THE EVALUATION OF ASSOCIATIVE ^-FIXATION IN BAHIA6RASS AND CORN BY ROBERT LARSON GREEN A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY JIVERSITY OF FLORIDA 1982 DEDICATED TO KAY AND MY PARENTS FOR THEIR LOVE AND KINDNESS ACKNOWLEDGMENTS I wish to express my deepest appreciation to Dr. Rex Smith, a man I have grown to admire during the past five years. His research genius and insight are unsurpassed. He is a kind man whose compassion is endless. Special thanks are given to Dr. Stanley Schank for his advice and help in many areas of my research program. The guidance and friendship of Drs. Albert Dudeck, L. Curtis Hannah, and Thomas Humphreys are deeply appreciated and will not be forgotten. I wish to acknowledge Dr. Donald Graetz, William Pothier, and Candy Cantlin for their help in soil nitrogen determinations. Special thanks are given to Dr. Ramon Littell for his friendship and statistical consultation. Thanks are given to James Milam and Leslie Villarreal for bacterial culture preparations. I wish to acknowledge the following people who provided help in my research program: Douglas Manning, Kenneth Cundiff, Keith Parsons, Loretta Tennant, Peter Mansanow, Peter Craig, Dale Bonne! 1, Matthew Shook, Hampton McRae, and Glen Weiser. Special consideration is given to Anthony Bouton. I wish to thank Edna Larrick who was so kind with her typing of this dissertation. This research program was supported by USAID Contract ta-c-1376. iii TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES ABSTRACT . . . CHAPTER I II INTRODUCTION EVALUATION OF N2-FIXATION IN BAHIAGRASS BY 15N-IS0T0PE DILUTION AND OTHER TECHNIQUES Introduction Materials and Methods . Results and Discussion III POTENTIAL FOR N^FIXATION IN ZEA MAYS GENOTYPES GROWN IN FLORIDA Introduction Materials and Methods . Results and Discussion IV CONCLUSIONS LITERATURE CITED . . BIOGRAPHICAL SKETCH . Page iii v vi i 5 5 8 15 31 31 35 43 65 68 74 IV Table 1 2 3 10 11 12 13 14 LIST OF TABLES Twenty-one bahiagrass genotypes surveyed for N?-fixation potential Diazothrophs included in inoculation mixture Analysis of variance for acetylene reduction 15 activity, top N, top dry weight, top N, and top % N in bahiagrass Correlation coefficients of several Np-fixation parameters Overall mean estimates of acetylene reduction 15 activity, top N, top dry weight, top N, and top % N in bahiagrass Genotype comparisons of inoculation effect (IE) adjusted means Analysis of variance and inoculation means of N balance measurements 15 Plant-soil recovery of N-labeled fertilizer . Diploid/tetraploid means and contrast procedure for various N?-fixation parameters Corn genotypes surveyed for Np-fixation potential Genotypes and experiment parameters of test-tube experiments Analysis of variance of first coring date, field experiment 1 Analysis of variance of second coring date, field experiment 1 Overall means by coring date of acetylene reduction activity, root dry weight, and soil moisture, field experiment 1 ... Page 9 11 16 17 18 21 23 27 28 36 41 44 44 45 LIST OF TABLES (Continued) Table Page 15 Analysis of variance for the combined data of field experiment 1 46 16 Analysis of variance for soil mineral content from coring date 1, field experiment 1 (part 1). . 48 17 Analysis of variance for soil mineral content from coring date 1, field experiment 1 (part 2). . 49 18 Analysis of variance for soil mineral content from coring date 1, field experiment 1 (part 3). . 50 19 Analysis of variance for soil mineral content from coring date 2, field experiment 1 (part 1). . 51 20 Analysis of variance for soil mineral content from coring date 2, field experiment 1 (part 2). . 52 21 Analysis of variance for soil mineral content from coring date 2, field experiment 1 (part 3). . 53 22 Overall mean soil mineral content of field experiment 1 (part 1) 54 23 Overall mean soil mineral content of field experiment 1 (part 2) 55 24 Analysis of variance of coring dates 1 and 2, field experiment 2 57 25 Analysis of variance for the combined data of field experiment 2 57 26 Overall means by coring date of acetylene reduction activity, root dry weight, and soil moisture, field experiment 2 58 27 Analysis of variance for greenhouse experiments 1 and 2 59 28 Overall means of acetylene reduction activity by sample date, top dry weight, root dry weight, and soil moisture from greenhouse experiments 1 and 2 60 29 Analysis of variance and genotype comparisons of test-tube experiments 62 30 Genotype means of acetylene reduction activity in n mole C?H4 evolved/ (core -h) for field and greenhouse experiments 63 vi Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE EVALUATION OF ASSOCIATIVE ^-FIXATION IN BAHIAGRASS AND CORN By Robert Larson Green December 1982 Chairman: Rex L. Smith Major Department: Agronomy Inoculation experiments were conducted on 21 Paspalum notatum Flugge genotypes and 15 Zea mays L. genotypes grown under field, green- house, and axenic conditions, using several N^-fixing bacteria. This study sought to (1) compare acetylene reduction activity (ARA) , yield, 15 N balance, and N-isotope dilution methods for estimating N^-fixation; (2) develop quick methods for screening genotypes with superior ability to support N?-fixation; and (3) evaluate the potential of bahiagrass and corn genotypes to support N^-fixation. Methods for estimating N„-fixation gave different values of fixed N. Percentage of total N in bahiagrass tops from N,,-fixation due to inoculation was estimated to be 8% from yield differences and 2% by 1 5 the N-isotope dilution method. Estimates of N^-fixation from ARA measurements were highly variable and inconsistent with those of other methods because control plants had higher nitrogenase activity than inoculated plants. Total amount of fixed N in inoculated and control vn plant-soil systems was calculated by the N balance method. An overall mean of 0.094 g N in excess of all N inputs was observed which indi- cated 46% of the N in these systems was fixed. Similar calculations 15 determined by ARA measurements and the N-isotope dilution method were 12% and 6%, respectively. Nitrogen balance determinations indicated that 81% of fixed N was located in root-zone soil. Corn genotypes did not respond consistently under field, green- house, and axenic conditions as determined by ARA. However, two geno- types, Asgrow RX-112 and Funk G-4864 had high ARA under most growing conditions. An overall significant response to inoculation for top growth and top N was observed in bahiagrass. However, no exceptional genotype could be identified. Tetraploid bahiagrass genotypes exhibited greater nitrogenase activity but were not as efficient as diploids for taking up fixed N and preventing its loss. Overall, bahiagrass genotypes assim- ilated 40% of 15N-labeled fertilizer while 43% of it was lost from plant-soil system. vm CHAPTER I INTRODUCTION Plant requirements for N exceed those of any other essential element for growth. Nitrogen is abundant in the atmosphere but in order to be available for plant growth it must go through a reduction process (fixation). The current input of fixed N in the world is about 116 x 105 MT of which 31% is fixed by industrial processes (Hardy and Havel ka, 1975). Biological fixation and the depletion of soil N account for the remainder. Petroleum resources supply the large quantity of energy needed for industrial fixation. Energy supplies are limited and expen- sive and their availability is unreliable. This concern coupled with the fact that the world's food production capability has failed to keep pace with the world's population has prompted additional interest in biological N^-fixation. Legumes play a major role in the world's food production. The legume-Rhizobium symbiosis, supplies essentially all the N that is required for production of these crops. However, on a worldwide basis, cereal and forage crops are grown on 90% more acreage than legumes (F.A.O., 1977). If biological N2~fixation in grasses could supply agron- omically important amounts of N, i.e., 15-30 kg N/(ha*growing season), its impact on world food production would be of great significance. Interest in N?-fixation in grasses was revised with the discovery that the diazotroph, Spirillum lipoferum (now described as Azospi rill urn 1 brazil ense and AzospiriHum lipoferum), was associated with the roots of the tropical grass Digitaria decumbens (Dobereiner and Day, 1976; Day et al., 1975b). Prior to these studies nitrogenase activity was observed in the rhizosphere of Paspalum notatum Flugge (Dobereiner et al., 1973; Dobereiner et al., 1972). Since then many studies have described asso- ciations between Np-fixing bacteria and numerous grass species including many of the world's most important grasses such as corn, rice, and wheat. Yield responses to the inoculation of grasses with these N,,- fixing bacteria have been successful but erratic. In addition to fixing N, these diazotrophs produce plant growth hormones which could be involved in yield responses (Tien et al., 1979; Gaskins et al . , 1977). Methods of estimating N?-fixation other than yield differences between inoculated and control plants are needed to give a more direct estimate of ^-fixation. Almost all the literature of ^-fixation deals with the measure- ment of acetylene reduction activity (ARA). Nitrogenase is able to reduce C?H2 to CJ\. so measurement of this reduction was proposed as an indirect method to assay for [^-fixation (Hardy et al., 1973). The preincubated washed root modification of this method, introduced and used by Dobereiner et al. (1972) has been criticized because of nonlinear, over-estimated activity and long lag periods (van Berkum and Bohlool, 1980). The major problems with the ARA method are high variability and measurements which are short term and indirect. Also the theoretical assumption that for every three moles of C^H. produced one mole of N2 will be fixed has to be verified by 15N incorporation. Estimates of ^-fixation by ARA often fail to support yield increases or correlate with inoculation treatments. However, this may not be a methodology problem but may indicate something about ^-fixation and yield responses. 15 Other methods of estimating ^-fixation include N balance and N-isotope dilution studies. These methods give more direct estimates of N2-fixation than those based on ARA but they are not without problems. Large amounts of initial soil N in N balance studies create difficulty in estimating fixed N because the small values of N fixed by the associative system are within the limits of experimental error of N measurements. Growing plants in soil or medium low in N may avoid the above problem. Studies using the N-isotope dilution method, i.e., methods using N fer- 1 C tilizer enriched with N, require an accurate estimate of soil N mineral- ization either by analytical methods or by using nonfixing control plants having similar N uptake patterns. Unfortunately, the establishment and maintenance of control plants in the nonfixing state for very long is difficult. The mineralization control used in analytical methods must be maintained at conditions similar to the root zone and must give slow uni- form changes in 14N : N ratios of available N over the experimental period. These results are difficult to accomplish. Studies utilizing 15 N fertilizer that is enriched with N provide estimates of N loss from the plant-soil system. These losses are from denitrificatin, volatiliza- tion, and leaching and must be considered in N balance studies. Also, estimates of percent recovery of fertilizer N by the plant can be calcu- lated. The most direct and definitive method of estimating ^-fixation 15 15 is by the incorporation of N from N2> This method is laborious, expensive, and difficult to employ in field studies. However, this method must be used to verify measurements from other methods. Each of the previously described methods for estimating N2~ fixation has its own merit and limitations. Concurrent utilization of several methods is advisable if not necessary. However, the final desired result of N?-fixation should be yield increases in the field or reduction of N fertilizer consumption. Field experimentation is expensive, time consuming, and progress can be slow. Quick screening methods to reduce the number of genotypes tested in the field would be a great asset to a breeding program. There- fore, quick screening methods under greenhouse or growth chamber condi- tions should be developed to select genotypes for field evaluation of N?-fixation potential. Very little research has been conducted in this area. Development of successful quick screening methods will depend on three factors. First, an accurate and consistent method of measuring N?-fixation must be available. Secondly, genetic variation within a species for the ability to support N?- fixation must exist for selection to be possible. Differences among genotypes, primarily determined by ARA, have been described within several species, including bahiagrass and corn. Lastely, genotype responses in the screening system must corre- late with field responses. The objectives of this study were 15 (1) to compare ARA, yield, N balance, and N-isotope dilution methods for estimating N?-fixation; (2) to develop quick methods to screen genotypes for the ability to support N^-fixation; and (3) to evaluate the potential of bahiagrass and corn genotypes to support N?-fixation. CHAPTER II EVALUATION OF ^-FIXATION IN BAHIAGRASS BY 15N-ISOTOPE DILUTION AND OTHER TECHNIQUES Introduction In most environments, nitrogen is the plant-growth-limiting resource. Energy to commercially fix N is expensive and limited. This concern, coupled with world food demand, has prompted recent interest in the associative symbiosis between N,,-fixing bacteria and the roots of grasses. It has been observed that these systems fix low but agronom- ical ly important amounts of N (Blue, 1974; Neyra and Db'bereiner, 1977; Day et al . , 1975a). Recent reviews on Np-fixation were written by Neyra and Db'bereiner (1977) and van Berkum and Bohlool (1980). Significant increases in plant growth as a result of inoculation with N?-fixing bacteria were reported with Pennisetum americanum (L. ) K. Shum. (Bouton et al., 1979), Cynodon dactyl on L. Pers. (Baltensperger et al. , 1978) , Pancium maximum Jacq. (Smith et al . , 1978), and Zea mays L. (Kapulnik et al . , 1981; Nur et al., 1980). However, a null response was reported with Pennisetum glaucum (L.) R. Br. (Barber et al . , 1979) and Z. mays (Albrecht et al . , 1981). Schank et al. (1979) reported a negative response with P_. americanum. At best, yield responses to inoculation have been erratic and unpredictable. Studies have shown genetic variation for N^-fixation in P_. ameri - canum (Bouton, 1977), Z_. mays (von Blilow and Dbbereiner, 1975), and P. 5 notatum (Benzion and Quesenberry, 1978). Tetraploid ecotypes of P.. notatum have been observed to exhibit greater diazotroph colonization, acetylene reduction activity, and greater amounts of photosynthate released from its roots than diploid ecotypes (Dbbereiner and Campelo, 1971; Db'bereiner et al., 1972; Vietor, 1982). Other studies with Z. mays (Albrecht et al., 1981) and C. dactyl on (Baltensperger et al., 1978) have indicated no significant differences among genotypes for the ability to support N^-fixation. Acetylene reduction activity (ARA) estimates of ^-fixation are often variable and fail to correlate with inoculum treatments (Albrecht et al., 1981); Weiser, 1980; Schank et al., 1979; Brown, 1976; Taylor, 1979; Smith and Schank, 1981). ^-fixing bacteria produce plant growth hormones (Tien et al., 1979; Barea and Brown, 1974) which may promote root development (Tien et al., 1979; Schank et al., 1979; Umali-Garcia et al., 1978; Schank et al . , 1981) and improve growth of young plants when measurable increases of ARA are not produced (Barea and Brown, 1974; Brown, 1976). Plant growth substances and N„-fixation may both be respon- sible for reported stimulation of growth. More definitive techniques such as N incorporation and N balance studies are needed for proper understanding of plant-bacterial association. Incorporation of N into plant tissue from N2 is considered the most direct and definitive evidence for ^-fixation. Yoshida and Yoneyama (1980) estimated N-- fixation rates of 1.4 mg N/(plant-13 d) with Oryza sativa L. Japonica var. ' Koshihikari ' . They calculated 20% of the fixed N was assimilated by the plant. Other grasses that have 1 5 been reported to assimilate I^L include JP. notatum and Digitana decumbens Stent. (De-Polli et all, 1977), 0. sativa (Ito et al . , 1980) and Saccharum officinarum L. (Ruschel et al . , 1975). However, Matsui 15 et al. (1981) did not detect enrichment of N in field-grown S. officinarum. 15., . Estimations of N^-fixation using N-isotope dilution and 'A' value calculations require a healthy nonfixing control plant or an accur- ate estimate of soil N mineralization by analytical methods. Rennie (1980) 15 found no significant difference between the yield difference and N- isotope dilution method for calculating percentage fixed N in Z. mays. His calculations were achievable because plants were grown axenically in vermiculite, i.e. (1) a plant-growth medium system of which total plant- available N inputs were very low and defined so determinations of fixed N were not confounded, and (2) control plants did not exhibit ARA. Previous greenhouse and field studies in our lab indicate (1) control plants exhibit ARA and (2) N assimilated from soil sources greatly exceeds assimilated fixed N (Bouton, 1977). Thus, the amount of fixed N which is relatively small is within the limits of experimental error of N measurements. The objectives of this study were to (1) screen a 21-genotype collection of bahiagrass for the ability to support ^-fixation; (2) examine ploidy effects upon Np-fixation potential in bahiagrass; 15 (3) compare ARA, yield, N balance, and N-isotopic dilution methods for estimating N?-fixation; and (4) estimate percent N fertilizer loss from the plant-soil system and percent N fertilizer recovery of the plant. Materials and Methods Twenty-one bahiagrass clones, which included 12 tetraploid and nine diploid genotypes, were obtained, courtesy of G. W. Burton, 21 Dec. 1979 (Table 1). These clones were maintained under a high fertility regime to promote rapid growth and provide adequate propagation material for the greenhouse study on 15 Mar. 1981. Ten uniform sprigs of each clone were planted, one each in a 15.2 cm plastic pot containing approx- imately 2,700 g pasteurized (24 h at 107°C and at a pressure of 6.9x10 Pa) growing median "soil" consisting of 95% coarse builders sand and 5% bentonite clay (V/V), pH = 8.4, and 17 ppm total N. A tygon tube (15.24 cm long, 1.27 cm diam) was inserted into each pot to facilitate subsurface irrigation and fertilization. The tube had emitting holes along its side and its bottom end was plugged with a stopper. It was believed that a dry medium surface would reduce diazo- trophic contamination by providing a nonsuitable environment for bacter- ial growth. In order to minimize light exposure of soil and prevent the growth of blue-green algae, a 1.9 cm layer of gravel was placed on the soil sur- face and the top of each pot was covered with aluminum foil, leaving only the plant exposed. Pot drainage holes were Closed with duct tape and small holes (0.5 mm) were punched to permit drainage. Plants were watered as needed but not enough to cause drainage. However, pots were flushed on a monthly basis to reduce salt accumulation. Research geneticist, AR, SEA, USDA, and the University of Georgia College of Agricultural Experiments Stations, Coastal Plain Station, Agronomy Department, Tifton, GA 31793. Table 1. Twenty-one bahiagrass genotypes surveyed for ^-fixation potential Genotype no. Tifton genotype ID Ploidy 1 Materia I ino.- 43 — kow no. 27 hi ant 66 2 26 17 64 3 33 29 63 4 11 30 59 5 14 10 59 6 14 10 58 7 38 4 59 8 10 18 53 9 10 18 54 10 15 22 48 11 15 22 47 12 49 13 46 — Tiftor hybrid no. — 13 78- 17 14 78-24 15 78- 16 16 78- 23 17 78- 17 18 15- 21 19 17- 29 20 78- 13 21 Pensacolar 4n 4n 4n 4n 4n 4n 4n 4n 4n 4n 4n 4n 2n 2n 2n 2n 2n 2n 2n 2n 2n Acquired from established sod, Plant Science Lab, University of Florida, Gainesville, FL. 10 Plants of each clone were graded according to size and sorted into five matched pairs; one of each pair was used for inoculation and the other was used for a control. The pairs were arranged in blocks and randomized according to a 21 x 5 type IV balanced incomplete block design (Cochran and Cox, 1957). One-liter cultures for each of seven diazotrophs (Table 2) were prepared in aerated liquid succinate N free media (SNF) described by Tyler et al . (1979) with trypticase (Baltimore Biological Laboratories) and (NH.)2S04 were included at 1.0 and 0.5 g/L, respectively. Cultures Q were grown for 18 h at 35°C to approximately 10 cells/ml. Cells were washed twice, centrifuged at 2,500 rpm for 10 min and resuspended in distilled water. Cultures were obtained from J. R. Milam, Department of Microbiology and Cell Science, University of Florida, Gainesville, Florida. Inoculated plants received 33 ml of a mixture containing equal culture volumes of each diazotroph. Control plants received the same amount of the above mixture after autoclaving. Plants were inoculated as above 6, 23, and 95 d after planting. Plants were fertilized biweekly with a complete N free Hoagland's solution (Hoagland, 1950). Total amount applied was equivalent to 26 and 70 kg/ha, P and K, respectively. Nitrogen was applied monthly in the form of KN03 (2.23 atom % 15N). A total of 0.064 g N/pot (a rate of 35 kg N/ha) was applied in four applications. Fertilization was omitted one month prior to harvest. 11 ro rO 03 03 n3 ro rd 03 -a -o "O "O -a -a ■o ■a S- S- t- i- S_ S- 5- i- o o o o o o O o - o 03 •!- 3 +J CD o| M| M| 12 Acetylene Reduction Plants were assayed for nitrogenase activity by acetylene reduc- tion activity (ARA) 122 to 142 d after planting. Pots were placed in 10 L plexiglass cylinders (51 cm deep, 16.5 cm diam). A top with a fitted rubber gasket and a sampling port (1.3 cm diam) for insertion of a septum was bolted to the cylinder. Prior to fastening, vasoline was applied sparingly to the gasket to create an air-tight seal. Cylinders were flushed with argon for 1 min through the sampling port and sealed with a septum. Oxygen concentration in the cylinders after flushing was approx- imately 6%. Acetylene was added through the septum to an approximate cone of 10% in cylinder gas phase. Cylinders were incubated 16 to 21 h in a growth chamber maintained at 30°C. Ethylene evolution was measured by gas chromatography. Yield, Total N, Exchangeable N, and N-determi nations 15, Plant tops (leaves and stolons) were harvested 156 d after plant- ing, dried at 60°C, and weighed. Soil from each pot was collected and the five replications from each genotype-inoculation treatment was pooled. Remaining plant parts were transplanted in the field for further evalua- tion. Plant tissue was ground through a 1 mm screen in a Wiley mill. In a procedure described by Weiser (1980), 0.2 g samples were block digested (Gallaher et al., 1975) in a mixture containing 1.5 g K2S04, 1.5 ml HgS04 solution (12 H2S04:88 H20:10 HgO V/V/W), and 6 ml cone HLSO.. Ammonium of the digestate was recovered by alkaline distilla- tion, collected in boric acid- indicator solution, and titrated with standard acid as described by Bremmer and Edwards (1965), procedure (A), 13 Ethanol was distilled between sample distillations to remove traces of + 4 NH. from the distillation apparatus. The distillates were acidified with 2 ml 0. 083 _N H^SO,, condensed to approximately 3 ml, and analyzed 15 for atom % N with a VG Micromass 602 E mass spectrometer. 15 Total N and atom % N of soil were determined by the above procedure except 4.0 g of screened soil was block digested in a mixture containing 3.2 g K2S04:CuS04 (9 : 1 W/W) catalyst, 6 ml cone H2S04 and 2 ml 30% H202 (Gallaher et al . , 1976). Exchangeable N from soil was extracted in KC1 , steam distilled and titrated. Screened, 30 g soil samples were shaken for 1 h in 100 ml 1 N KC1 and allowed to settle 24 h. Determination of NH» and N0~ followed a steam-distillation- titration procedure using MgO and Devarda alloy described by Bremner (1965), except 75 ml aliquots were distilled. Nitrogen Balance and Accumulation Nitrogen accumulating in the plant-soil system (PSS) in excess of all N inputs was calculated as follows: [final plant nitrogen (Plant N) + final soil nitrogen (Soil N)] - [(initial sprig nitrogen) + (initial soil nitrogen) + (fertilizer nitrogen)]. Fertilizer and initial soil contributed 0.064 and 0.046 g N/pot, respectively. Total N was deter- mined for representative sprigs of each genotype. Overall mean N content was 0.004 g/sprig. Plant dry weights were estimated to compute plant N with the ratio 1 top dry weight : 1.64 plant dry weight. This ratio was determined from top, root, and plant dry weights of a similar study (Benzion, 1978). 15, l-Labeled Fertilizer Recovery 15, 14 The amount of N-labeled fertilizer remaining in the PSS was calculated with equations of Hauck and Bremner (1976): 15 (1) grams N-labeled fertilizer remaining in the soil = (soil N)(atom % 15N soil - 0.3663) . d (atom % 15N fert. - 0.3663 (2) grams N-labeled fertilizer taken up by plant = (plant N)(atom % 15N plant - 0.3663) (atom % 15N fert. - 0.3663) The above calculations relate to the currently accepted value of 0.3663 1 5 atom % N for naturally occurring N. The percent fertilizer N recovered by the plant and PSS can be calculated with the following equations: (3) percent fertilizer N recovered by plant = equa tion (2 0.064 g N (2.23 atom % N) (4) percent fertilizer N recovered by PSS = yg — x (100); and equations (1) + (2) 0.064 g N (2.23 atom % N) T5~ x (100) 15 Results and Discussion ARA, Top 15N, and Top Yield An overall significant response to inoculation was observed for most measured variables (Table 3). Overall, genotypes responded simi- larly to inoculation, indicated by the absence of significant genotype x inoculation interaction (Gxl). A highly significant genotypic effect 15 was observed for all measured variables, except top N. Acetylene reduction activity (ARA) was highly variable which was demonstrated by a C.V. =70.0 and an activity range of 30 to 4420 n mole C?H»/ (cylinder «h). This N?-fixation estimate was inconsistent with inoculum treatments and negatively correlated with top dry weight (Table 4). Initial sprigs were soil grown and could not be sterilized so they were the probable source of unknown diazotrophs. Because of this and other contamination sources, control plants exhibited higher ARA rates than inoculated plants (Table 5). Other possible sources of contamination include transfer from inoculated plants, airborne organ- isms, and incompletely pasteurized soil. Possibly, bacterial populations of control plants grew from an initial low population to one comparable to inoculated plants at the time of ARA assays. Tops of inoculated plants had a significantly (P = 0.09) greater 15 N-isotope dilution than control plants indicating greater fixation in the former (Table 5). If inoculated plants are considered fixing systems (fs) and control plants nonfixing systems (nfs), then the principle of 15 N-isotope dilution may be used to calculate the amount of N that was fixed due to inoculation (Rennie, 1979): re 0) ■!- QJ fO >— -Q >■> +-> c o (1) +J U CO +-> i— CD J (]> S- CD CD > a o fO u CI) i s- 7?. ld s- CD N +J c CD • i— u +J ■- I. 0) CI J D- <+- CD * r-. * «* * i— * en * C\J * CM * LT) * en * 1— * CD +J N l/l ■r- >i to +-> S- • t— CD o >. in >, i- S- 0) a> +-> > +-> > c o £Z o 0J o ID o u cu O CD Table 5. Overall mean estimates of acetylene reduction 15 activity, top N, top dry weight, top N, and top % N in bahiagrass. Measured variable Adjusted inoculation means t Inoc (+) Control (-) - n mole C :2H4/(cylinder.h) — ARA 623 784 % 15N Top 15Nt" 1.6159 1.6430 Top dry weight 5.73 5.40 Top N 0.024 0.022 01 |VI Top %N 0.447 0.432 'Adjusted means are required for comparisons between treatments in balanced incomplete block designs. tEach plant received 0.064 g N in the form of KN0., (2.23 atom % 15N). 19 15 percentage N? fixed , atom % N excess (fs) , - I - T"r X I atom % N excess (nfs) 00. Calculations indicated 2% of top N was from fixation due to inocula- tion. Top dry weight (top DW) and top N responses to inoculation were 15 relatively more pronounced than responses of top N-isotope dilution (Table 5). Genotype adjusted means of top DW and top N ranged from 3.27 to 9.25 g and 0.015 to 0.035 g, respectively. Inoculated plants had 8% more top N than control plants. Estimates of percent top N from fixation due to inoculation 15 differ when calculations are based on top N and top N-dilution measure- ments. The former estimates, based on yield measurements and N determin- 15 ations are subject to greater experimental error than N-isotope dilution measurements, which only require a representative sample of the test crop for determination (Rennie and Rennie, 1973; Rennie et al . , 1978). Thus an estimate of 2% top N from ^-fixation due to inoculation, as determined by 15N-isotope dilution may be more precise than an estimate of 8% based on differences in yield. Top percent N was unaffected by inoculation (Tables 3 and 5). Significant differences of top percent N among genotypes were probably related to differences in yield. This is indicated by a significant negative correlation (r = -0.41; P = 0.0001) between top percent N and top DW. Inoculation effect (IE) of each pair was computed (IE = inoculated plant minus control plant) and analyzed to identify genotypes with a 20 superior response to inoculation. All overall genotypic effect on IE was not significant for any measured variable (P = 0.80). This indi- cated that genotypes responded uniformly to inoculation for all measured variables. Genotype comparisons were made on adjusted IE means by L.S.D. procedures (Table 6). A significance level of 0.01 was chosen because a nonsignificant Gxl was observed earlier in Table 3 and a significant difference among genotypes for IE represents Gxl. Therefore, conserva- tive comparisons between genotypes are justified and preferred. No genotype had a unique IE for any measured variable. This observation further demonstrates that basically, genotypes responded to inoculation uniformly. Differences in the sign of IE suggest Gxl but this was not significant as shown in Table 3. An overall response to inoculation was significant but minimal and genotypes responded similarly to inoculation. Thus, no exceptional genotype could be identified by its inoculation response. Consequently, identification of superior genotypes for N,,-fixation potential by inocu- lation response may be difficult. Caution must be used when ^-fixation estimates are based on ARA, top growth, or both. Verification of sup- rior genotypes and quantification of N,,-fixation rates should include 15N-data. Nitrogen Balance and Accumulation Because only the top growth was harvested, a top dry weight to plant dry weight ratio of 1 to 1.64 was used to estimate plant DW. This ratio was determined from top, root, and plant dry weights of Table 6. Genotype comparisons of i 21 noculation effect (IE) adjusted means. IE (inoculated - control ) adjusted means for four variabl esT Genotype 1 no. ARA 15 Top IDN Tof ) dry wei ght Top N n mole C2H4/(cyl inder • h)- -atom % 15N grams — 1 159 -0.0325 -0.50 -0.001 2 -62 -0.0442 0.31 0.002 3 -116 0.0855 1.64 0.004 4 -351 0.0145 0.19 0.001 5 135 -0.0830 -0.55 -0.002 6 78 0.0646 0.51 0.002 7 -36 0.0535 -0.35 0.000 8 -461 -0.0316 0.32 0.002 9 226 -0.0166 -0.26 0.000 10 -246 0.0098 0.49 0.001 11 -218 0.0443 -0.40 0.000 12 1185 -0.0056 1.68 0.003 13 -296 -0.0483 -0.71 -0.003 14 807 0.0566 0.07 0.000 15 130 -0.0680 0.07 0.001 16 143 0.0213 -0.46 -0.001 17 -351 0.0704 -0.54 -0.004 18 -238 0.0261 0.69 0.003 19 116 -0.0573 0.46 0.001 20 189 0.0697 -0.48 -0.002 21 32 -0.1540 1.32 0.004 'Adjusted means are required for comparisons between treatments in balanced incomplete block designs. Ifleans within each column are not different at the P = 0.01 level as determined by pair wise t test. 22 a similar study (Benzion, 1978). This ratio is reasonable because of the similarity of Benzion 's study with this one, i.e. (1) it was a 130 d greenhouse study surveying three tetraploid and three diploid bahiagrass genotypes for N^-fixation potential, (2) plants were grown in 15.3 cm clay pots containing a loamy sand which was amended with N at a cali- brated rate of 60 kg N/ha, and (3) plants were inoculated with A. brasi Tense strain JM 125 A2 or A. pas pal i. Further, it is reasonable to calculate plant N with top percent N rather than plant percent N because Benzion was unable to detect a difference between top and root percent N. He reported mean values of 0.46 and 0.50% N, respectively. These values are comparable to the mean top N value of 0.43% determined in this study. Some error in the estimation of plant N is not serious because root N was estimated to only contribute about 7% of plant-soil system N (Table 7) An analysis of variance was made on plant N but its purpose was to produce adjusted means of this parameter for N balance equations. Analysis of variance for soil percent N and soil N were not appropriate because soil samples were pooled; however, analysis of variance was made to produce adjusted means. Mean soil N was 63 ppm and ranged from 47 to 91 ppm N. Genotype adjusted means of soil N ranged from 0.128 to 0.246 g/pot. Differences among genotypes for soil N may be related to differences in root system development and mineral uptake capability which could (1) influence the amount of N loss from the plant-soil system and (2) influence the quantity of organic matter present in the plant-soil system. The first will be discussed in conjunction with plant recovery of N-labeled fertilizer and the second is supported by several observations: (1) fine root matter was observed in soil samples 23 Table 7. Analysis of variance and inoculation means of N balance measurements. Source Analysis of variance df Plant N MS F Accumulated N/PSS MS F 20 0.0006 8.87 Block 20 0.0001 Genotype Inoculation 1 0.0004 7.08 Genotype x Inoculation 20 0.0001 0.74 Adjusted inoculation means 0.0001 0.0085 136.45 0.0011 0.0016 Inoc. (+) Control (-) 17.53 25.72 Plant N gran 0.040 IS IN 0.037 Soil N 0.170 0.169 Accumulated N/plant-so system il 0.096 0.092 *,**Significant at the P = 0.05 and 0.01 levels, respectively. ""Adjusted means are required for comparison between treatments in balanced incomplete block designs. 24 following screening; (2) there was a significant correlation between soil N and top DW or top N (Table 4); and (3) soil N was 63 ppm of which 15 ppm was extractable NHl N (only a trace amount of N0~ N was detected). There- fore 76% of soil N was organic and in either plant or microbial tissue. Soil N accounted for 81% of the accumulated N (N accumulated in the plant-soil system in excess of all N inputs). The above was indi- cated by a correlation coefficient of 0.95 between soil N and accumulated N (Table 5). Yoshida and Yoneyama (1980) observed that 75 to 81% of total fixed N in rice plant-soil systems was located in root zone soil. A significant response to genotype and inoculation was observed for accumulated N (Table 7). Genotype adjusted means ranged from 0.054 to 0.192 g/PSS. A significant Gxl indicated genotypes did not respond uni- formly to inoculation. However, inoculated PSS tended to accumulate an average of 4% more N than controls. The accumulated N is probably from ^-fixation but a genotype's ability to utilize N and reduce its loss may relate to several factors such as (1) plant size, i.e., root system development, and (2) plant mineral uptake capability. Thus a genotype that has more fixed N in its PSS may be one that reduces N loss because of certain characteristics in its root system. The significant genotypic effect on accumulated N may 15 relate to the above and will be discussed with N-labeled fertilizer recovery. An adjusted grand mean of 0.094 g accumulated N/PSS during the 156 d duration of the experiment or 0.60 mg N accumulated/(PSS«d) may be accurate and indicates considerable amounts of fixed N in inoculated and control PSS. An adjusted grand mean for ARA was 703 n mole C^^/ 25 (cylinder* h). This activity can be converted to fixed N/(PSS*d) by assuming that 3 moles of C?H» are produced per 1 mole N„ fixed. This calculation indicates 0.16 mg fixed N/(PSS*d). Estimates of amounts of fixed N by ARA and accumulated N may be extrapolated to (g N/(ha«d) by multiplying them by 553,191 , a factor which equates the area of a pot to a hectare. Acetylene reduction activity and accumulated N measure- ments extrapolate to 90 and 330 g fixed N/(ha«d). It is difficult to explain why these estimates differ so much but they are independent methods of measurement and ARA depends on short term measurements unlike those of N balance. Measurements of N in the entire PSS are important because they account for N which is unaccounted for with measurements of plant yield. Approximately 81% of accumulated N was located in root-zone soil. There- fore hL-fixation may have its largest impact on root-zone soil so it is important to include this part of the PSS in N?-fixation studies. Soil N which was 76% organic should eventually become available for plant growth. 15 N-labeled Fertilizer Recovery 15, Assumptions for use of N are discussed by Hauck and Bremner (1976), two of which are important for the measurements of this study: (1 ) complex elements (those containing two or more isotopes) in the 15 natural state have a constant isotope composition (0.3663 atom % N); and (2) the distribution of applied N among different plant parts or within a plant part is uniform. 15, Analysis of variance was made on N-parameters so that adjusted means could be obtained. Overall adjusted means for plant and soil 26 15 recovery of N-labeled fertilizer were 0.026 and 0.010 g, respectively (Table 8). Genotype adjusted means for the above ranged from 0.017 to 0.040 g recovered by the plant and 0.006 to 0.017 g recovered in the 15 soil. Adjusted means for percent N-labeled fertilizer recovered by the PSS and plant were 56.9 and 40.4%, respectively. Genotype adjusted means for the above ranged from 39.9 to 79.4% for the PSS and 26.8 to 62.4% for the plant. 15 Approximately 72% of the total recovered N-labeled fertilizer 15 was located in the plant. Plant recovery rates of N-labeled fertilizer were probably influenced by genotype. Plant factors that promote root system development and mineral uptake would enhance fertilizer recovery rates and reduce fertilizer N losses. It follows that a genotype with enhanced fertilizer uptake would have enhanced uptake of N from other 15 sources such as biologically fixed N. Genotypes with high N-labeled fertilizer recovery would be genotypes with reduced losses of fixed N and therefore have higher N accumulation values (accumulated N in excess of all N inputs). The above is demonstrated by significant correlations 15 between accumulated N and percent N-labeled fertilizer recovered by plant and PSS (Table 4). A genotype with superior N efficiency may be a genotype that supports N^-fixation by leaking photosynthate from its roots (Vietor, 1982), but it may also be a genotype that has enhanced N uptake abilities that allow it to take up fixed and soil N that otherwise may be lost. The above supposition can be supported by comparing tetraploid and diploid means of several measured variables (Table 9). Tetraploid genotypes exhibited greater nitrogenase activity, i.e., support Np-fixation Table 8. Plant-soil recovery of 15 N-labeled fertilizer. 27 N-source Adjusted means' Inoculated Control Overall 15 grams N-labeled fertilizer recovered* Plant 0.026 Soil 0.010 Plant-soil system 0.036 Plant Soil 15 Plant 41.5 Soil 15.7 Plant-soil system 57.0 0.025 o.on 0.036 percent recovery of N-labeled fertilizer 39.3 17.2 56.7 15 atom % N excess 1.2499 0.1148 1.2771 0.1233 0.026 0.010 0.036 40.4 16.4 56.9 1.2635 0.1195 Adjusted means are required for comparison between treatments in balanced incomplete block designs. X 15 TEach plant-soil system received 0.064 g N-labeled fertilizer. 28 CJ> n csj o O o CO o o o CM o o CM o o o o o o o o o o o 3 o o o o o o o -1 +1 +1 +1 +] +1 +1 +1 ■> o CM CM o LCI U3 o "3- o o ^3- co CM o o o o o o o o o D O o o o c; ■a- o o cr, o o o o o o o o D O o c: o o o o H +1 +1 + +1 +i +1 +1 r en M l£) M O CM o a- oc 1— CO o o o o us CM o o o <= o o o o S- OJ >> 0) r— O 0> ,— S- .— QJ QJ QJ > XI IT OJ f— u *— 29 better than diploids as observed by ARA. Vietor (1982) observed that the 14 1 percentage of the C-labeled photosynthate appearing in the root wash solution was significantly greater for tetraploids than diploids. Reports of greater diazotroph colonization and greater ARA in association with tetraploids compared with diploid roots (Dobereiner and Campelo, 1971; Dbbereiner et al . , 1972) are consistent with Vietor's data and the data of 15 this study. Measurements of N-isotope dilution of top growth indicate equal amounts of fixed N in diploid than tetraploid plants. However, 15 measurements of N-isotope dilution of soil indicated there is 26% more fixed N in the soil of diploids and tetraploids. Diploid PSS had 54% 15 more fixed N. Diploid genotypes recovered 17% more N-labeled fertilizer. 15 Plant-soil systems of tetraploid plants exhibited 9% more N-labeled fertilizer loss. The above and previous data indicate that tetraploid genotypes exhibit greater nitrogenase activity, but are not as efficient as diploids for taking up fixed N and therefore experience greater losses of fixed N from the plant-soil system. Diploids retain greater amounts of N in their root- zone soil. Two characteristics may be important in the selection for N efficiency: (1) the ability to support diazotrophic colonization; and (2) the ability to take up fixed N and not lose it. The interrelation- 15 ship between these characteristics may become clearer with N^-incorpor- ation studies in conjunction with ARA, N uptake, and N accumulation mea- surements. Medium that is nearly N free should be used so as not to confound N accumulation measurements. Inoculation did not affect fertilizer recovery rates (Table 8). 15 However, soil of inoculated PSS exhibited a greater N-isotope dilution. 30 If inoculated PSS are considered a fixing system and control PSS nonfix- 15 ing, then calculations of N-isotope dilution can be made to estimate the percentage soil N from fixation due to inoculation. Seven percent of soil N was from fixation due to inoculation. Earlier calculations estimated 2% of top N was from fixation due to inoculation. The amount of fixed N due to inoculation in the PSS can be estimated by adding 2% of the plant N to 7% of the soil N. Calculations indicate 0.013 g N are fixed due to inoculation. The above calculation does not estimate total amounts of fixed N in the PSS because these calculations consid- ered control PSS as nonfixing while in reality they exhibited N„- fixation rates nearly equal to inoculated rates. 15 An average of 43% of N-labeled fertilizer was lost. Renme 1 5 (1979) reported losses of N-labeled fertilizer of 22%. Possible sources of loss include denitrification, volatilization, or leaching. If the assumption that no significant isotope discrimination occurs 15 14 during these loss processes and thus the N/ N ratio remains constant, then the isotope dilution procedure will remain unaffected. However, measurements based on total N are underestimated due to losses. In summary, N„-fixation contributed considerable amounts of N to the bahiagrass PSS (46% of PSS N was fixed). However, inoculation of these systems with N^-fixing bacteria accounted for only 8 and 2% of top N, as estimated by differences between inoculated and control plants in yield and N-isotope dilution measurements, respectively. Thus, estimates of N„-fixation which were based on differences between inoc- 15 ulated and control plants in yield and N-isotope dilution measurements account for only a portion of the total fixed N within the bahiagrass PSS. CHAPTER III POTENTIAL FOR ^-FIXATION IN ZEA MAYS GENOTYPES GROWN IN FLORIDA Introduction The purpose of this introduction is to supplement the literature that has been previously cited with more information concerning Np-fixa- tion in Zea mays L. The reader is referred to the other introductions in this dissertation for a more complete review of associative N?-fixation. Reports of N?-fixing bacteria living in association with the roots of 1. mays have been published (O'Hara et al . , 1981; von B'ulow and Db'ber- einer, 1975). Nitrogen fixing bacteria may be attracted to root exudates of several grasses including 1. mays (Alvares-Morales and Lemos-Pastrana, 1980). Following inoculation they have been observed adsorbing to root surfaces and found in the middle lamellae of root cells within one week (Umali-Garcia et al., 1978). The potential of these associations involv- ing important cereal crops and forage grasses is significant because economically important amounts of nitrogen could be fixed biologically. The actual contribution of N2~fixing bacteria to the N economy of Z. mays and other grasses remains elusive. Efforts to further understand this association are currently under investigation. Significant yield increases in 1. mays due to inoculation with N?-fixing bacteria have been reported (Kapulnik et al . , 1981; Nur et al . , 1980; O'Hara et al., 1981; Cohen et al . , 1980; Rennie, 1980). However, 31 32 nonsignificant responses in yield of 1. mays inoculated with N^-fixing bacteria have been reported (Albrecht et al., 1981; Albrecht et al., 1977). Inoculation has been observed to increase the diazotrophic popu- lation in the rhizosphere without concurrent increases in nitrogenase activity as estimated by acetylene reduction activity (O'Hara et al . , 1981). Other reports indicate increased acetylene reduction activity (ARA) in inoculated roots but without apparent N benefit to the plants (Okon et al., 1977; Barber et al., 1976). This suggests that other factors such as bacterial-produced plant growth hormones are at least partially responsible for yield responses because diazotrophs have been reported to produce plant growth hormones (Gaskins et al., 1977; Tien et al., 1979). Measurement of ARA is the most common method for estimating N„- fixation. This technique is easily employed in field, greenhouse, and growth chamber studies. Unfortunately it is an indirect method of short duration which is extremely variable. Several factors influence ARA which include soil type or field location (von Bulow and Dobereiner, 1975; Cohen et al., 1980) climatic conditions such as air temperature and light (Cohen et al . , 1980; Albrecht et al., 1977; Weier, 1980), soil moisture levels (Weier, 1980; Smith and Schank, 1981; Smith et al . , 1982), reduced N levels in the growth medium or soil (Barber et al . , 1979; Day et al . , 1975b; Cohen et al . , 1980; Smith et al . , 1977), 02 and C02 levels of assay vials or root-soil cores (Dobereiner et al . , 1972; Smith et al., 1982) plant maturity (Albrecht et al . , 1981; von Bulow and Dobereiner, 1975) and seasonal variations (Smith et al . , 1982; Weier et al., 1981). 33 In addition, ARA is highly variable even when the above factors are held constant. Corroboration of ARA with other methods of estimating N?- 15 fixation such as N balance and N-isotope dilution should be considered. Tests for the ability to support PL-fixation which are conducted under field conditions are closest to natural conditions, but are time- consuming and expensive. Development of quick methods, adapted to screen- ing seedlings or young plants under greenhouse or growth chamber condi- tions would be helpful by predicting field results and permitting the simultaneous evaluation of large numbers of plant genotype-bacterial strain combinations for associative N?-fixation. Superior genotypes selected by these methods would be evaluated further under field conditions. Inoculation studies have been conducted using several methods for growing plants in growth chambers under initially axenic conditions. They include growing plants in vials containing vermiculite (Rennie, 1980), in bottles containing sand (Albrecht et al., 1981), in test tubes containing Fahraeus nitrogen and carbon free medium (Schank and Smith, 1980), in test tubes or "Whirl Pac" plastic bags containing autoclaved soil (Schank et al., 1979), and in Leonard's bottle-jar assemblies con- taining washed sand, soil, or loess (Cohen et al . , 1980) or an acid washed sand-vermiculite mixture (Rennie and Larson, 1979). In addition to these systems, several have been developed for inoculation studies under greenhouse conditions. They include growing plants in large ceramic containers containing soil (Bouton et al . , 1979), in plastic or mason jar assemblies containing soil (R. L. Smith and S. C. Schank, unpublished data, 1981), and sand-vermiculite (O'Hara et al . , 1981) or sand-soil pot cultures (Albrecht et al . , 1981). 34 Studies under greenhouse or growth chamber conditions usually have been unsuccessful in repeating field responses. Bouton (1977) was unable to repeat yield responses observed in the field with Pennisetum americanum L. ' Gahi-3' inoculated with Spiri Ilium lipoferum in nonaxenic tests. Schank et al . (1979) grew genotypes of £. americanum that had been selected because of their high and low response to field inoculation in test tubes and plastic bags containing autoclaved soil. Yield responses were negatively correlated with inocualtion and it was con- cluded that the system did not provide a suitable environment for screen- ing bacterial-plant associations and predicting their response in the field. Schank and Smith (1980) grew corn and millet plants axenically in test tubes containing nitrogen and carbon-free Fahraeus media and inoculated with Azospi rill urn. Acetylene reduction activity was high and bacteria colonized root surfaces but no consistency was observed with previous field inoculation responses. However, latter axenic test-tube and field experiments indicated that several corn and millet genotypes with high ARA in test tubes also had high ARA under field conditions (Schank and Smith, unpublished data, 1981). They also found that ARA under test-tube conditions was influenced by several factors including agar concentration, mineral salt composition, and inoculum strain. They felt that the latter may represent bacterial strain-plant cultivar complementation. There is a need for further development and refinement of quick screening methods. In order for these methods to be successful, an accurate method for measuring N„-fixation must be available and genetic variation within a species for the ability to support [^-fixation must 35 exist. Variations among 1. mays genotypes for ARA have been described byvon Blilow and Db'bereiner (1975). The objectives of this study were to (1) develop quick methods to screen corn genotypes for the ability to support associative N?- fixation; (2) compare these methods to core acetylene reduction assays of field-grown plants; and (3) evaluate the potential of corn genotypes to support N?-fixation. Materials and Methods A collection of 26 Zea mays genotypes (Table 10) were grown under field, greenhouse, and growth chamber conditions. Plants were grown in 2.2 L plastic jar assemblies in a greenhouse and axenically in test tubes in a growth chamber. Field Experiments Two experiments were conducted in the field at the University of Florida, Institute of Food and Agric. Sci . , Beef Research Unit (B.R.U.), Gainesville, Florida. The first experiment tested the response of 15 corn genotypes to inoculated and indigenous N^-fixing bacteria. The second experiment tested the response of these same genotypes to indig- enous N?-fixing bacteria. Previous studies at this location indicated abundant populations of undefined N?-fixing bacteria. Soil at this location is primarily a Wauchula fine sand, a poorly drained hyperthermic siliceous Ultic Haplaquod, which was analyzed at the termination of the first experiment and found to be pH = 7.22, 0.70 ppm NH» N, 0.10 ppmNO'N, and 1.16% organic matter. In both experiments, fertilizer was applied 36 «a-ocMLnoooor^c\j CO i— I O I 00 CO CM I i — D1 l/) u o o o (1) cu C o O en c» Q. O. CD CT c i — c f= Ei c c i— •1— o o •r- ■r- ro -r- -r- C\l CO +J < ) CD LO CIJ 1 "3 5 X X3 CO •I- -r- O •i- •!- CL CO i— co ro cm OJ OJ CU Q- QJ 37 prior to planting at a rate of 20 kg P and 70 kg K/ha, and 4.48 kg/ha of fritted trace elements (FTE 503; 5 B, 5 Cu, 29 Fe, 12 Mn, 0.3 Mo, 11 Zn in g/100 kg). Nitrogen was applied as NH.NO- after seedling emer- gence at a rate of 25 kg/ha. Planting. In the first field experiment 15 genotypes were planted in a strip-plot design, replicated six times. Each block con- sisted of 15 rows (one row for each genotype) that were 91 cm apart. Each row (main plot) was divided into two 6.1 m subplots for inoculation treatments. Subplots within each rowwere separated by 3 m to minimize cross inoculation. Each subplot had 20 hills in each of which three seeds were planted. The second field experiment had 16 genotypes planted in a random- ized complete block design, replicated four times. Each block consisted of 16 rows (one row for each genotype) 91 cm apart and 6.1 m long. Twenty hills were planted in each row as before. Field experiments were weeded and irrigated when needed. Lanate was applied several times for insect control. Inoculation. Liquid cultures of S125 and S145 (two Azospi rill urn spp. , isolated by R. L. Smith, Univ. of Florida; S125 was isolated from the B.R.U. ) were grown in 10 L batches in a fermenter to an approximate o cone of 10 cells/ml. Both diazotrophs were grown in aerated succinate N free medium with trypticase and (NH.KSO, added, as described in Chapter II. The appropriate subplots of the first experiment were inoculated two and ten days after planting with a mixture containing equal volumes 38 of each diazotroph at an approximate rate of 19 ml/m at the first inoculation and one half that for the second. Sprinkling cans were used to apply the inoculum. Control subplots received autoclaved inoculum at the same rate as specified above. Sprinkler irrigation was applied immediately after application. Acetylene reduction activity, root dry weight, soil moisture and mineral analysis. Sampling for ARA involved taking 18 cm soil -root cores in metal tubes either 7.5 or 10.2 cm diam and 36 cm long. One end was sealed with a rubber cap clamped in place and the other end with a rubber septum. Both experiments were sampled twice, the first experiment 59 to 70 d and 83 to 93 d after planting and the second 77 to 84 d and 100 to 107 d after planting. After taking a core containing a corn plant cut off at ground level, each tube was capped, then flushed with argon for 1 min, followed by the addition of acetylene to approximately 10% of the gas phase volume. Tubes were incubated at 30°C in a growth chamber for 18 to 24 h and assayed for ethylene evolution by gas chro- matography. Following ARA measurements root-soil masses were removed from each metal tube. Soil samples were taken for analyses and oven- dried to determine moisture content. Soil analyses for P, K, Ca, Al , Cu, Fe, Mg, Mn, Zn, NH» N, H0~ N, pH, and percent organic matter were conducted on four replications by the IFAS Soil Science Laboratories. Each root mass was washed over a 16-mesh metal screen, dried at 60°C, and weighed. 39 Greenhouse Experiments Two greenhouse experiments were conducted to (1) reevaluate Z. mays genotypes grown in the field under conditions intermediate to field and axem'c studies and (2) test a second group of genotypes in an effort to find greater genetic variation for the potential to support N?-fixation (Table 10). In both experiments, plants were grown in 2.2 L plastic jar assemblies (Inmark Inc., Atlanta, Georgia) with drip irriga- tion and gravity drainage. Each jar was filled with soil from the B.R.U. that had been screened through a 16-mesh wire screen. A 3.8 cm hole through which plants could emerge was punched into each jar lid. A sec- ond hole, 1.3 cm diam, was drilled in the bottom of each jar for drainage. In order to minimize light exposure of soil and prevent the growth of blue-green algae each jug was sprayed with silver reflective paint and a 2 cm layer of waxed sand was placed on the soil surface. Planting. In the first experiment, 14 genotypes were planted (three seeds per jar) and then thinned by selecting one uniform seed- ling. Jars were arranged in a complete block design with five replica- tions. In the second experiment, 13 genotypes were planted and arranged in the above manner with four replications. In both experiments, con- trol jars consisted of a complete jar assembly without a plant. Gen- eral maintenance of irrigation and insect pest control with Lanate were used as needed. Plants of the first experiment showed signs of mineral deficiency (not N), approximately 43 d after planting so they were fertilized with a complete N free Hoagland's solution at a rate of 1.2 kg P and 12.6 kg K per hectare. 40 Acetylene reduction activity and yield measurement. Acetylene reduction activity procedure involved placing each jar assembly into 10 L plexiglass cylinders as described in Chapter II. Cylinders were flushed with argon for 1 min, followed by the addition of acetylene to an approximate cone of 10% in cylinder gas phase. Cylinders were incu- bated for 18 to 24 h and assayed for ethylene evoluation by gas chroma- tography. Both experiments were analyzed for ARA twice, the first experiment 45 to 52 d and 67 to 74 d after planting and the second experiment 30 to 35 d and 60 to 65 d after planting. Following ARA measurements, tops were harvested, dried at 60°C and weighed. Root-soil mass was removed from each jar and root dry weight and soil moisture measurements were made as. described on page 38. Test-Tube Experiments Sixteen 1. mays genotypes previously grown under field and greenhouse conditions were tested in four growth chamber experiments (Table 10). A description of each experiment can be seen in Table 11. Seedlings were grown axenically, according to a procedure described by Schank and Smith (1980), in 15x200 mm test tubes containing 60 ml of nitrogen and carbon free, semi-solid (6 g agar/L) Fahraeus medium (Fahraeus, 1957) with 4 ml/L of 0.5% bromothymol blue solution (dis- solved in 10% ethanol as a pH indicator. The medium was adjusted with NaOH to pH = 7. In experiments 3 and 4 succinate (1.965 g/L) was included for a carbon source to stimulate bacterial growth because in experiments 1 and 2 low ARA was encountered with diazotrophs grown on medium without an added carbon source. Active bacterial populations should not be the limiting factor in this screening method and 41 <3- +j c CD E s- cu Q. X LU co co +-> c +-> CI J c F 0) E S- Cl) s- n 0) x a. fl) X LU (1) x> 3 4-> 4-> in Q) +-> 4- O in <- CI) CM +-> cu 4-> E= E (T1 Q. C" X Cl) LU E s- C!) O- X CD ■n c: . +-> o +-> c C CI) CD e) E s- cu Q. X LU CD -Q fO i— CM lO CQ 00 00 CM O CO LO O r— LO CO O Q +-> • re a: u • •r- CJ Q. ro CD lO ID M O) O O CM CO 00 i— O I i— i— 00 CO CM VD UD CM en 1-^ O O 00 i— CO o • ■r- a •4-> • (O C£ o • •I- <_> a. ro CD E Q CO o • cr -o o; CI E • ca c_) +j o: — - m u CU ■1 — +-> C1J Q. Ci) Q- S- E LO CD CM CO c 0) >- • — ■-o •i— in c 01 •r— !- rC -£-> r— c c X 0) 0) CM F •c— «=C| 42 differences in ARA should reflect plant contributions (root exudates released into the medium) to the plant-bacteria association. Seeds were surface sterilized by consecutive dips in 95% ethanol for 30 s, two dips in 50% Clorox for 5 min each, and 3% hydrogen peroxide for 2 min. After thorough washing in sterile water, seeds were germi- nated on either agar plates (0.70% agar with 0.50% sucrose added to test for bacterial contamination) or sterile paper towels. Poor and uneven germination with the former method was experienced so paper towels were used. The towel method resolved the germination problem but seeds were not tested for bacterial contaimi nation. However, seedlings that were germinated on paper towels and then transferred to test tubes appeared to have a low contamination rate. Any contaimi nated tubes were discarded. Two days after seeds were sterilized test tubes were inoculated 8 9 with a suspension of washed cells (approximately 10 to 10 cell/ml) of Azospi rill urn brasilense strain JM 125 A2 or a mixture containing equal volumes of the above diazotroph and Azospi rill urn brasilense strain CD (ATTCC 29729). Each test tube received either 0.2 or 1.0 ml of inoculum mixed in the warm (40 to 45°C) melted growth medium which then was allowed to solidify to its semi-solid state. Cultures of diazotrophs for inoculation were grown and cells washed as described in Chapter II. Seedlings approximately 2 cm in length were transferred to test tubes 3 to 4 d after seeds were sterilized if germinated on paper towels and 3 to 8 d if germinated on agar plates. Seedlings were placed on stainless steel screens inserted into each test tube and suspended 0.5 cm above the medium in order to prevent leaching nutrients of the seed 43 from influencing bacterial growth and ARA. Control tubes without seed- lings were included in the experiment and the analysis of variance to determine the effect of a plant on ARA. Plants were grown in a growth chamber maintained at 30°C with a 14 h day. Light intensity was 149 pE/(m2's) at plant level. Acetylene reduction activity was measured 15 to 20 d after seedlings were transferred by adding 10% acetylene to each tube and incubating for 20 to 24 h in the growth chamber maintained as above. Ethylene evolution was measured by gas chromatography. Results and Discussion Field Experiments Field Experiment 1. On the first coring date neither inoculation or genotype significantly affected ARA, root dry weight (Root DW), or soil moisture (Table 12). Analysis of variance from the second coring date indicated the same as above, except a genotype x inoculation inter- action was significant (Table 13). Overall means of these measurements can be seen in Table 14. Acetylene reduction activity, Root DW and soil moisture measurements were highly variable as indicated by high coeffi- cients of variation. Combined analyses of both coring dates indicated ARA was not significantly different on each coring date but soil moisture was sig- nificantly higher and root dry weight significantly lower on the second coring date (Tables 14 and 15). The latter observation may relate to the fact that plants were beginning to show signs of senescence on the second coring date. Root dry weights were significantly different among genotypes which indicated differences in growth characteristics. 44 Table 12. Analysis of variance of first coring date, field experiment 1 df ARA Root MS DW F Soil Mois MS 106.70 ture Source MS F F Rep 5 199282 T 756 Genotype 14 17550 0.99 313 1.74 0.77 1.37 Rep x genotype 70 17714 180 0.56 Inoculation 1 16207 18841 0.86 236 1.81 0.01 0.02 Rep x inoculation 5 131 0.18 Genotype x inoc ulation 14 19309 1.07 163 1.62 0.29 1.20 :,**Significant at the P = 0.05 and 0.01 levels, respectively. Table 13. Analysis of variance of second coring date, field experiment 1 df ARA Root MS DW F Soil Mois MS ture Source MS1 F F Rep 4 409,767 610 288.61 Genotype 14 74,163 1.09 185 1.69 0.97 0.98 Rep x genotype 56 68,141 109 0.99 Inoculation 1 17,539 0.23 243 2.63 0.01 0.00 Rep x inoculation 4 75,789 94 2.11 Genotype x inoculation 14 117,957 1.21 34 0.86 0.74 * 2.21 :, ^Significant at the P = 0.05 and 0.01 levels, respectively. +f r— T- o M- 45 O +1 *3- o CO CO 46 ■s r— i— CO r- -r- >, 47 Correlations of combined data indicated ARA increased with increasing levels of soil moisture (r = 0.29, P = 0. 0001) but was uncorre- cted with inoculation or Root DW. Weier et al. (1981) and Smith et al (1982) reported similar correlations between ARA and soil moisture. Root dry weight decreased with increasing levels of soil moisture (r=-0.20, P = 0.0003), but other factors, such as maturity may have caused this phenomenon. More coring dates during the experiment are needed to further investigate correlations between ARA, Root DW, and soil moisture. Mineral analyses taken from soil samples of the first coring date indicated inoculated plants were not significantly different from control plants for all tested elements (Tables 16 - 18). Mineral anal- yses taken from the second coring date indicated the same as above except inoculated plants had significantly higher percent organic matter (Tables 19-21). Genotypes were significantly different for soil percent organic matter on the first coring date (Table 18). Soil samples from inoculated plants had higher overall mean soil mineral concentrations than controls in 79% of the comparisons in Tables 22 and 23. However, when data from each coring date were combined, only NH. N was significantly correlated with inoculation (r = 0.20, P = 0.002). Soil mineral concentrations of several elements were significantly correlated with soil moisture, pH, and Root DW. Soil mineral concen- trations for all tested elements pH and organic matter were uncorre- cted with ARA. Smith et al. (1982) found that soil Zn content was positively correlated and calcium soil content negatively correlated with ARA. 48 5 t <£> 00 •— ro o o o U3 <* 3 r-» co •vt" CO o CTl CO ro OJ < 4- O c 5 ro CO "3 C CTl o c ) O v£> LO co c ) ^J- co t£> LO o Lf ) CM C\J r-- r~ OJ c ) r— CM •^t C\ J CM c 5 r^ 1 c > "~ Ln CO CO CO hs 1 — 1 — c ) OJ <— O o o c ) o o <3- r^ CO LO r-» CM o c\ <- l _ ir> ^1 lo o «* lo o a ) CO co r-^ 1 c\ J o a i o LO CT •l 03 O c > 1 cr> CM WD C o r^. o^ CO r^ c «* co ■ — 1 — LO c ) «* CTl IX) LO CO <3 <3- r^ CO c ) CO 1 ro >* CM ro ■* ** o c: o ro u O) +j o Q- fO c >> +-> E 3 o C o X c o a> (1) ■T c: ai Q- CD re •i- Q- >> >i 4-> X — X 4-> 49 o o O CD Q. >-o ID CO a) «* u-> ID ■ — r— LO I— O CT> 00 ■* «3" CO (O CD > E (T3 O) E t- 53 o Q «± i— O O O O i— CO O O o o 54 55 CM o 8 ° o CO o o 56 Field Experiment 2. A significant genotypic effect was not observed for ARA, Root DW or soil moisture for either coring date (Table 24). Data from both coring dates were combined and analyzed and indicated ARA and soil moisture were significantly higher on the second coring date (Tables 25 and 26). As observed in the first field experiment, ARA was positively correlated with soil moisture (r = 0.60, P= 0.0001). Soil moisture content was significantly different among genotypes when the data were combined. In summary, corn genotypes grown under field conditions and inoculated with N^-fixing bacteria or grown in the presence of indig- enous N2-fixing bacteria, were not significantly different for ARA. These measurements were highly variable but significantly correlated to soil moisture. High variation for ARA in field experiments exists but the failure of these experiments to identify genotypes for high ARA may not relate entirely to methodology. Insufficient genetic variation may have been present in the corn hybrids. Greenhouse Experiments Significant differences among genotypes were not observed for ARA in either greenhouse experiment (Table 27). Activities were not different for the first and second acetylene reduction assay within both experiments (Table 28). These measurements were highly variable as indicated by a high C.V. Top dry weights were significantly differ- ent among genotypes in experiment 1 and root dry weights were signif- icantly different among genotypes in experiments 1 and 2. In experi- ment 1 Top DW exceeded Root DW and in experiment 2 the opposite was observed. Differences among genotypes for Top DW and Root DW indicated 57 Table 24. Analysis of variance of coring dates 1 and 2, field experiment 2. Source Rep Genotype Rep Genotype df 3 15 3 15 ARA MS Root DW MS f coring date 1 262,580 9,109 0.93 cori ng date 2 263,514 25,480 1.63 362 112 0.73 354 70 0.89 Soil Moisture MS F 54.00 0.47 0.85 6.87 1.11 0.95 f,**Significant at the P = 0.05 and 0.01 levels, respectively. Table 25. Analysis of variance for the combined data of field experiment 2. df ARA Root MS DW F Soil Mc MS )isture Source MS F F Rep 3 246,800 399 37.00 Genotype 15 16,634 1.18 124 1.45 0.93 * 1.13 Rep x genotype 45 14,068 85 0.82 Date 1 701 ,446 ** 25.00 22 0.17 244.00 ** 116.89 Date x genotype 15 17,954 57 0.42 r, ^Significant at the P = 0.05 and 0.01 levels, respectively. 53 Table 26. Overall means by coring date of acetylene reduction activity, root dry weight, and soil moisture, field experiment 2. Parameter Mean Range C.V. ARA 1 ARA 2 129 ± 18 277 ± 22 n mole C?H4/(core*h) 3.0 - 613.0 50.0 - 819.0 grams 114 62 Root DW 1 15.05 ± 2.09 Root DW 2 14.01 ± l.li 2.66 - 55.78 2.64 - 65.11 79 67 percent Soil moisture 1 8.07 ± 0.22 5.36 - 11.70 22 Soil moisture 2 10.84 ± 0.14 6.91 - 13.16 10 59 60 Table 28. Overall means of acetylene reduction activity by sample date, top dry weight, root dry weight, and soil moisture from greenhouse experiments 1 and 2, Greenhouse experiment 1 4. Parameter Mean Range1 C.V. ARA 1 387 ±36 - 74 to 1,188 76 ARA 2 320 ± 28 -125 to 974 71 Top DW 4.92 t 0.11 3.53 to 7.06 1.70 + 0.06 0.64 to 3.11 19 Root DW 30 Soil Moisture 16.94 : t 0.39 3.96 to 20.04 19 Gree ihouse experiment 2 n mole C2H4/ (cylinder 'h) ARA 1 270 ± 26 - 6 to 857 69 ARA 2 244 ± 20 - 19 to 553 58 Top DW 1.34 t 0.06 0.70 to 3.38 29 Root DW 3.02 t 0.12 0.93 to 4.88 28 Soil Moisture 18.07 ± 3.39 4.78 to 21.66 15 Negative ARA values were not measured but obtained when ARA of control jars exceeded sample jars. Control jars did not contain a plant. 61 they had different growth characteristics. Soil moisture was sig- nificantly different among genotypes in experiment 2, but was uncorre- cted with ARA of the second sample date. However, soil moisture was correlated (r = 0.28, P = 0.02) with ARA of the second sample date in experiment 1. In summary, corn genotypes grown under greenhouse condi- tions were not significantly different for ARA. These measurements were highly variable but only in a few cases did controls (no plant) exceed plant systems. Test-Tube Experiments Acetylene reduction activity was significantly different among genotypes in experiments 1 to 3 but not in experiment 4 (Table 29). In these analyses, controls (no plant) were considered a "genotype." In experiments 2 to 4, controls exhibited as high or higher ARA than plant systems. This indicates that the addition of a plant to the system does not contribute to ARA activity and some genotypes signif- icantly decreased it. One explanation for this is that seedlings compete with bacterial populations for growth producing substances, i.e., mineral salts and possibly carbon sources. Asgrow RX-112 exhib- ited high ARA in experiments 2 and 3, while ARA of other genotypes was inconsistent. The addition of succinate as an energy source and use of greater amounts of inoculum enhanced ARA up to 2,729%. Comparisons of Screening Methods Genotype ARA from greenhouse and test-tube experiments was compared to field activity to determine if field performance could be predicted by using greenhouse and/or test-tube results (Table 30). 62 <£ CO CO CD OD i- ID in C^ O CT> C\J CM r--ounc\jOLr)iD(XJix> 3 3 Ol Ol Ol CO N i— m co i— CO r-~ O CO <*«roo en •a- i— en CO Ch o < > m i^ *j- >* H '7 <"1 ^ CO, S- i— CM c -a: o o •r- r^ uo co C3 QJ ja ■= 3 *e < «t CQ CQ CO CO CO C0r.-ur>cr>r-.u3OC0 cocmcocoiOf— r-^o co to ro w QJ i/i •r- QJ . ■a e -a ■r- QJ O QJ 3 C Q.,— X U qj qj c id c'^ . "*" <" CNJ o o o O CO cm in ro r^. r- i co i— o > ■— CU QJ "O QJ CO C_J CC Ll_ O o i- io ^ CO CO GO CO CO CO o?3 M- QJ QJ O CM i — \£> r^ r^ lt> ld ^- r- O VO r— S- CM r- S- i— JT Q_ O O -C Q> ro UH C QJ QJ o C C ° :ec?£ £ oj£ 2 "X ° a- c in -r- QJ QJ 3 g_ 63 «a- O VD o en co co •.- «n cm O D JD ll. o a z rv ^ n «^ cri CO lt> no co CO in ro CM CM i— >— i— i— ro cm co co CD OJ "O O CD 3 S- S- c cr 3 .* cu Lorocr>covDfOco^J-i — locmi— *d un lH co ro co ^r*3- 0. 10). However, more liberal observations may be warranted and useful in screening studies. In general, genotypes responded erratically to all growing con- ditions. Asgrow RX-112 responded well in field experiment 1, test-tube experiments, and greenhouse experiment 1, but poorly in field experi- ment 2. Funk G-4864 responded well in both field experiments but poorly in greenhouse experiment 2 and on an average in test-tube exper- iment 1. From the limited data in this study a correlation probably does not exist between a genotype's ARA under axenic, greenhouse, and field conditions. However, a few genotypes may show promise for consistently having high ARA under axenic, greenhouse, and field conditions. Asgrow RX-112 and Funk G-4864 are two genotypes that should be further tested. CHAPTER IV CONCLUSIONS Methods of estimating N„-fixation gave different values of fixed N in a bahiagrass study under greenhouse growing conditions. Percentage of total N in bahiagrass tops due to N^-fixation from inoculation was 8% when estimated by yield differences between inoculated and control plants 1 5 and 2% when estimated by the N-isotope dilution method. Acetylene reduction activity (ARA) was highly variable and negatively correlated with the above methods of estimating N^-fixation. Estimates of N?-fixation due to inoculation did not measure total fixation because control plants exhibited fixation rates nearly equal to inoculated plants. However, measurements of accumulated N in excess of all N inputs were measurements of total fixation, i.e., measurements from the N balance method. These measurements in inoculated and control plant- soil systems indicated 46% of the N in the system was fixed. Thus N,,- fixation estimates which were based on differences between inoculated and control plants underestimated and accounted for only a portion of total N?- fixation. Estimates of total N„-fixation were calculated from ARA and N-isotope dilution measurements and indicated thatl2%and 6% of the N in the plant-soil system was fixed. These total N^-fixation estimates wer lower than estimates from the N balance method. However, the three esti- mates are from independent measurements because (1) ARA depends on short- term measurements, and (2) N-isotope dilution estimates considered 65 66 control plant-soil systems as nonfixing. Estimates of hL-fixation from N balance calculations may be closest to actual fixation rates because of the above limitations of the other two methods. These calculations indicated that agronomically important amounts of N were fixed. It is important to remember that even when inoculation responses are low, significant PL-fixation could be occurring. The N-balance method of estimating hL-fixation indicated that 81% of the N in excess of all N inputs was located in the root-zone soil. Therefore, N?-fixation may have its largest impact on root-zone soil so it is important to include this part of the plant-soil system in N?-fixation studies. Seventy-six percent of the soil N was in the organic form and was not available to plants. This N should be miner- alized over time and should become available for plant use. 15 The utilization of N-enriched fertilizer expanded the scope of the N balance study by permitting plant fertilizer N uptake and loss estimates from the plant-soil system. These measurements indicated 1 5 that 40% of the N-labeled fertilizer N was located in plant tissue 15 while 16% remained in the soil. Forty-three percent of the N-labeled- fertilizer N was lost. Since losses were not considered in the N„- fixation calculations, estimates of hL-fixation by the N balance method could be conservative. Development of quick screening procedures for the ability to support N?-fixationwere difficult to evaluate because corn genotypes were not significantly different for ARA in field and greenhouse exper- iments. Significant differences among genotypes were observed within test-tube experiments but genotype ARA between experiments was 67 generally inconsistent. Acetylene reduction activity had large amounts of extraneous variation but the difficulty in obtaining significant genotypic responses for ARA may partially relate to insufficient genetic variation in ability to support associative N~-fixation. However, more liberal observations indicate two genotypes, Asgrow RX-112 and Funk G-4864 show promise and should be tested further. A significant overall response to inoculation was observed in the bahiagrass study. However, genotypes responded similarly to inocu- lation and no exceptional genotype could be identified because the inoculation response measured was small and within the limits of exper- imental error. Tetraploid bahiagrass genotypes exhibited greater nitrogenase activity but were not as efficient as diploids in taking up fixed N. Therefore, two characteristics may be important in the selection of N efficiency: (1) the ability to support diazotropic colonization and (2) ability to take up fixed N and prevent its loss. This second characteristic may be related to several plant factors such as root development and mineral uptake efficiency. LITERATURE CITED Albrecht, A. L. , Y. Okon, and R. H. Burn's. 1977. Effects of light and temperature on the association between Zea mays and Spirillum lipo- ferum. Plant Physiol. 60:528-531. Albrecht, S. L. , Y. Okon, J. Lonnquist, and R. H. Burris. 1981. Nitro- gen fixation by corn-Azospirillum associations in a temperate cli- mate. Crop Sci. 21:301-306. Alvarez-Morales, R. A., and A. Lemos-Pastrana. 1980. Chemotaxis of Azospi ri 1 1 um-1 i poferum and Azospiri 11 um-brasi Tense towards root exudates of graminae. Rev. Latinoam. Microbiol. 22:131-136. Baltensperger, A. A., S. C. Schank, R. L. Smith, R. C. Littell, J. H. Bouton, and A. E. Dudeck. 1978. Effect of inoculation with Azospi rill urn and Azotobacter on turf bermudagrass genotypes. Crop Sci. 18:1043-1045. Barber, L. E. , S. A. Russell, and H. J. Evans. 1979. Inoculation of millet with Azospi rill urn. Plant and Soil 52:49-57. Barber, L. E. , J. D. Tjepkema, S. A. Russell, and H. J. Evans. 1976. Acetylene reduction (nitrogen fixation) associated with corn inoculated with Spirillum. Appl. Environ. Microbiol. 32:108-113. Barea, J. M. , and M. E. Brown. 1974. Effects on plant growth produced by Azotobacter paspali related to synthesis of plant growth regu- lating substances. J. Appl. Bacterid. 40:583-593. Benzion, G. 1978. Interaction of Paspalum notatum genotypes with associative Np-fixing bacteria. M. S. Thesis. Univ. Florida. Benzion, G. , and k. H. Quesenberry. 1978. Interaction of Paspalum notatum genotypes with associative N?-fixing bacteria. Agron. Abstr. 70:70. Blue, W. G. 1974. Efficiency of five nitrogen sources of Pensacola bahiagrass on Leon Fine Sand as affected by lime treatments. Soil and Crop Sci. Soc. Fla. Proc. 33:176-180. Bouton, J. H. 1977. Response of pearl millet inbreds and hybrids to inoculation with Spirillum li poferum. Ph.D. Dissertation. Univ. Florida. Univ. Microfilms. Ann Arbor, Mich. Diss. Abstr. Int. 38:5125-B. Order No. 7806676. 68 69 Bouton, J. H. , R. L. Smith, S. C. Schank, G. W. Burton, M. E. Tyler, R. C. Littell, R. N. Gallaher, and K. H. Quesenberry. 1979. Response of pearl millet inbreds and hybrids to inoculation with Azospi ri 1 1 urn- bras i lense. Crop Sci . 19:12-16. Bremner, J. M. 1965. Methods of soil analysis: Part 2. Chemical and microbiological properties. Agronomy 9:1195-1201. Bremner, J. M. , and A. P. Edwards. 1965. Determination and isotope- ratio analysis of different forms of nitrogen in soils: I. Apparatus and procedure for distillation and determination of ammonium. Soil Sci. Soc. Am. Proc. 29:341-348. Brown, M. E. , 1976. Role of Azotobacter paspali in association with Paspalum notatum. J. Appl. Bact. 40:341-348. Cochran, W. G. , and G. M. Cox. 1957. Experimental Designs. W. A. Shewhart and S. S. Wilks (ed.) John Wiley & Sons, Inc., New York, New York. Cohen, E. , Y. Okon, J. Kigel, I. Nur, and Y. Hem's. 1980. Increase in dry weight and total nitrogen content in Zea mays and Setaria italica associated with nitrogen-fixing Azospi rill urn spp. Plant Physiol. 66:746-749. Day, J. M. , D. Harris, P. J. Dart, and P. van Berkum. 1975a. The broadbalk experiment. An investigation of nitrogen gains from nonsymbiotic fixation, p. 71-84. Nitrogen fixation by free- living microorganisms. International Biological Programme Series , Vol. 6. Cambridge University Press, Cambridge, England. Day, J. M. , M. C. P. Neves, and J. Dbbereiner. 1975b. Nitrogenase activity on the roots of tropical forage grasses. Soil Biol. Biochem. 7:107-112. De-Pol 1 i , H. , E. Matsui , and J. Dbbereiner. 1977. Confirmation of nitrogen fixation in two tropical grasses by '5N? incorporation. Soil Biol. Biochem. 9:119-123. Db'bereiner, J., and A. B. Campelo. 1971. Nonsymbiotic nitrogen fixing bacteria in tropical soils. Plant and Soil- Spec. Vol. :457-470. Db'bereiner, J., J. M. Dart, and P. J. Dart. 1972. Nitrogenase activity and oxygen sensitivity of the Paspalum notatum-Azotobacter paspali association. J. Gen. Microbio. 71:103-116. Dbbereiner, J., and J. M. Day. 1976. Associative symbioses in tropical grasses: Characterization of microorganisms and dinitrogen-fixing sites, p. 518-538. In W. E. Newton and C. J. Nyman (ed.) Pro- ceedings of the First International Symposium on Nitrogen Fixation. Vol. 2. Washington State Univ. Press, Pullman, Washington. 70 Ddbereiner, J., J. M. Day, and P. J. Dart. 1973. Rhizosphere associa- tions between grasses and nitrogen-fixing bacteria: Effect of 0~ on nitrogenase activity in the rhizosphere of Paspalum notatum. Soil Biol. Biochem. 5:157-159. Fahraeus, G. 1957. The infection of clover root hairs by nodule bac- teria studied by a simple glass slide technique. J. Gen. Microbiol. 16:374-381. Food and Agricultural Organization of the United Nations. 1977. Annual fertilizer review for 1977. Food and Agriculture Organization of the United Nations, Rome. Gallaher, R. N. , C. 0. Weldon, and F. C. Boswell. 1976. A semiautomated procedure for total nitrogen in plant and soil samples. Soil Sci . Soc. Am. J. 40:887-889. Gallaher, R. N. , C. 0. Weldon, and J. G. Futral. block digester for plant and soil analysis. Proc. 39:803-806. 1975. An aluminum Soil Sci. Soc. Am. Gaskins, M. H. , M. Garcia, T. M. Tien, and D. H. Hubbell. 1977. N?-Fixation and growth substance production by Spirillum lipoferum and their effects on plant growth. Plant Physiol. 59:128. Hardy, R. W. F. , R. C. Burns, and R. D. Holsten. 1973. Applications of the acetylene-ethylene assay for measurement of nitrogen fixa- tion. Soil Biol. Biochem. 5:47-81. Hardy, R. W. F. , and U. D. Havelka. 1975. Nitrogen fixation research: A key to world food? Science. 188:633-643. Hauck, R. D. , and J. M. Bremner. 1976. Use of tracers for soil and fertilizer nitrogen research. Adv. Agron. 28:219-265. Hoagland, D. R. , and D. I. Arnon. 1950. The water-culture method of growing plants without soil, Calif. Agric. Exp. Stan. Circular 347. Berkeley, California. Ito, 0., D. Cabrera, and I Watanabe. 1980. Fixation of dinitrogen-15 associated with rice plants. Appl . Environ. Microbio. 39:554-558. Kapulnik, Y. , S. Sarig, I. Nur, Y. Okon, J. Kigel, and Y. Henis. 1981. Yield increases in summer cereal crops in Israeli fields inocu- lated with Az^sjD|riJJjum. Expl. Agric. 17:179-187. Matsui, E. , P. B. Vose, N. S. Rodrigues, and A. P. Ruschel. 1981. Use of '^N enriched gas to determine N2~fixation by undisturbed sugarcane plants in the field, p. 153-161. In P. B. Vose and A. P. Ruschel (ed. ) Associative N„- Fixation. CRC Press, Boca Raton, Florida. 71 Neyra, C. A., and J. Dobereiner, 1977. Nitrogen fixation in grasses. Adv. Agron. 29:1-38. Nur, I., Y. Okon, and Y. Hem's. 1980. An increase in nitrogen content of Setaria italica and Zea mays i nocul ated wi th Azospirill urn. Can. J. Microbiol. 26:482-485. O'Hara, G. W. , M. R. Davey, and J. A. Lucas. 1981. Effect of inocula- tion of Zea mays with Azospirill urn brasilense strains under temper- ate conditions. Can. J. Microbiol. 27:871-877. Okon, Y. , S. L. Albrecht, and R. H. Burn's. 1977. Methods for growing Spirillum lipoferum and for counting it in pure culture and in association with plants. Appl . Environ. Microbiol. 33:85-88. 15 Rennie, R. J. 1979. Comparison of N-aided methods for determining symbiotic di nitrogen fixation. Rev. Ecol . Biol. Sol. 16:455-463. 15 Rennie, R. J. 1980. N- isotope dilution as a measure of di nitrogen fixation by Azospirill urn brasilense associated with maize. Can. J. Bot. 58:21-24. Rennie, R. J., and R. I. Larson. 1979. Dim'trogen fixation associated with disomic chromosome substitution lines of spring wheat. An. J. Bot. 57:2771-2775. Rennie, R. J., and D. A. Rennie. 1973. Standard isotope versus nitro- gen balance criteria for assessing the efficiencies of nitrogen sources for barley. Can. J. Soil Sci . 53:73-77. 15 Rennie, R. J., D. A. Rennie, and M. Fried. 1978. Concepts of N usage in dim'trogen fixation studies, p. 107-133. In Isotopes in biological dim'trogen fixation. International Atomic Energy Agency, Vienna. Ruschel, A. P., Y. Hem's, and E. Salati. 1975. Nitrogen-15 tracing of N-fixation with soil-grown sugar cane seedlings. Soil Biol. Biochem. 7:181-182. Schank, S. C. , and R. L. Smith. 1980. Growth responses and acetylene reduction of several grass/bacteria associations grown in vitro. Can. J. Genet. Cytol . 21:677. Schank, S. C. , R. L. Smith, and G. C. Weiser. 1979. Responses of two pearl millets grown in vitro after inoculation with Azospirill urn brasilense. Soil Crop Sci. Soc. Fla. Proc. 39:112-115. Schank, S. C. , K. L. Weier, and I. C. MacRae. 1981. Plant yield and nitrogen content of a digitgrass in response to Azospirill urn inoculation. Appl. Environ. Microbio. 41:342-345. 72 Smith, R. L. , J. H. Bouton, S. C. Schank, and K. H. Quesenberry. 1977. Yield increases of tropical grain and forage grasses after inocu- lation with Sjrh^ilUim JJ_p_oferum in Florida, p. 307-311. In A. Ayanaba and P. J. Dart (ed.) Biological N2-fixation in farming systems of the tropics. John Wiley & Sons, Inc., New York, New York. Smith, R. L. , and S. C. Schank. 1981. Factors affecting nitrogenase activity of grass field plots, p. 495. In Current perspectives in nitrogen fixation. Australian Acad. Sci . Canberra. Smith, R. L. , S. C. Schank, J. H. Bouton, and K. H. Quesenberry. 1978. Yield increases of tropical grasses after inoculation with Spirillum lipoferum. Ecol . Bull. (Stockholm). 26:380-385. Smith, R. L., S. C. Schank, J. R. Milam, and R. C. Littell. 1982. Statewide search for highly active associative N^-fixation systems. (In press) Soil Crop Sci. Soc. Fla. Proc. 42. Taylor, R. W. 1979. Response of two grasses to inoculation with Azospirillum spp. in Bahamian soil. Trop. Agric. 56:361-366. Tien, T. M. , M. H. Gaskins, and D. H. Hubbell. 1979. Plant growth substances produced by Azospirillum brasilense and their effect on the growth of pearl millet (Pennisteum americanum L.) Appl . Environ. Microbio. 37: 1016-1024. Tyler, M. E. , J. R. Milam, R. L. Smith, S. C. Schank, and D. A. Zuberer. 1979. Isolation of Azospirillum from diverse geographic regions. Can. J. Microbiol. 25:693-697. Umali-Garcia, M. , D. H. Hubbell, and M. H. Gaskins. 1978. Process of infection of Panicum maximum by Spirillum lipoferum. Ecol. Bull. (Stockholm). 26:373-379. Van Berkum, P., and B. B. Bohlool. 1980. Evaluation of nitrogen fixa- tion by bacteria in association with roots of tropical grasses. Microbio. Rev. 44:491-517. 14 Vietor, D. M. 1982. Variation of C-labeled photosynthate recovery from roots and rooting media of warm season grasses. Crop Sci. 22:326-366. von Bulow, J. F. W. , and J. Dobereiner. 1975. Potential for nitrogen fixation in maize genotypes in Brazil. Proc. Natl. Acad. Sci. U.S.A. 72:2389-2393. Weier, K. L. 1980.. Nitrogenase activity associated with three tropical grasses growing in undisturbed soil cores. Soil Biol. Biochem. 12:131-136. 73 Weier, K. L. , I. C. Macrae, and J. Whittle. 1981. Seasonal variation in the nitrogenase activity of a Panicum maximum var. Trichoglume pasture and identification of associated bacteria. Plant and Soil. 63:189-197. Weiser, G. C. 1980. Evaluation of Panicum virgatum L. for nitrogen related agronomic characters. Ph.D. Dissertation. Univ. Florida. Univ. Microflims. Ann Arbor, Mich. Diss. Abstr. Int. 41:1603-B. Order No. 8025401. Yoshida, T. , and T. Yoneyama. 1980. Atmospheric dinitrogen fixation in the flooded rice rhizosphere as determined by the N-15 isotope technique. Soil Sci . Plant Nutr. 26:551-559. BIOGRAPHICAL SKETCH Robert Larson Green was born on November 7, 1951, in Miami, Florida, and graduated from South West Miami Senior High School in 1970. He received a Bachelor of Science degree in Biology from Florida State University in 1974 and a Bachelor of Science degree in Ornamental Horticulture in 1977. In 1979 he received a Master of Science degree in Horticultural Science from the University of Florida. In June 1979, he started a program of study and research at the University of Florida leading to the degree of Doctor of Philosophy with a major in agronomy and a minor in botany. 74 I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Rex L. Smith, Chairman Professor of Agronomy I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. J?£.LS? Albert E. Dudeck Associate Professor of Ornamental Horticulture I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. A e^fe-^ ( iVu-n n^A- Larkin C. Hannah Associate Professor of Vegetable Crops I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Thomas E. Humphreys Professor of Botany I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Stanley C. Schank Professor of Agronomy This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1982 X.& 9* iqri( Dean, College of Agriculture Dean for Graduate Studies and Research uilB^lI'-oHmA ifaifi