Full Text 2

Biol Fertil Soils (2004) 39:391–397 DOI 10.1007/s00374-004-0728-4

ORIGINAL PAPER

Hannan H. Youssef · Mohamed Fayez · Mohamed Monib · Nabil Hegazi

Gluconacetobacter diazotrophicus: a natural endophytic diazotroph of Nile Delta sugarcane capable of establishing an endophytic association with wheat Received: 18 March 2003 / Accepted: 30 December 2003 / Published online: 2 March 2004
Springer-Verlag 2004

Abstract Gluconacetobacter- like diazotrophs were encountered as dense populations inside the root and stem tissues of sugarcane cultivated in ancient agricultural fields of the Nile Delta. Counts of >105 cells g-1 were recorded in root and stem samples. The leaves contained a smaller population (<103 g-1). The typical dark-orange colonies which developed on LGIP agar plates were purified. Identification was performed with the API microtube systems: API 20E for Enterobacteriaceae and API 20NE for non-Enterobacteriaceae. API profiles of the local isolates were closely related to those of the type culture Gluconacetobacter diazotrophicus (ATTC 49037). The isolates successfully reduced C2H2 and produced appreciable amounts of ethylene in the presence of cane juice. This suggested that the local isolates are closely related to the type strain G. diazotrophicus. Wheat seedlings were inoculated with a number of isolates under gnotobiotic conditions. Both optical and scanning electron microscopy showed that endophytic Gluconacetobacter spp. were present in all the samples tested. They were observed in apparently intact and enlarged epidermal root cells, and also in stem tissues, indicating that the bacterium was able to migrate upward into the shoot tissues. Although Gluconacetobacter inoculation did not stimulate the growth of the cereal plant, the results obtained are particularly interesting because this bacterial species was capable of colonizing the internal tissues of wheat, not considered a natural host until now. Keywords Gluconacetobacter diazotrophicus · Diazotorophs · Associative nitrogen-fixers · Sugarcane · Wheat

Introduction A new diazotroph among the a-subclass of the protobacteria, Gluconacetobacter diazotrophicus , was originally isolated from surface-sterilized roots and stems of sugar cane using a semi-solid N-deficient medium with cane juice as the C source (Cavalcante and Dobereiner 1988; Gillis et al. 1989). This bacterium is of special interest because, besides fixing N2 in the presence of KNO3 and at low pH values <3.0 (Stephan et al. 1991), it can excrete almost half of the fixed N2 in a form potentially available to plants (Cojho et al. 1993). Such an endophytic diazotroph was also isolated from sugarcane in Mexico, Cuba and Australia (Li and MacRae 1991, 1992; FuentesRamirez et al. 1993). The occurrence of the microorganism was reported in roots, tubers and stems of sweet potato which confirm the endophytic nature of this particular diazotroph (Dobereiner et al. 1988). In several other instances, however, difficulties in finding this bacterium may be related to methods used for surface sterilization and isolation. Therefore, we report in the present paper the most successful procedures of sterilization and isolation for this bacterium, in addition to tracing the specific occurrence of the microorganism in Egyptian sugarcane. The ability of tested isolates to establish an endophytic association with wheat, as a non-legume, was also investigated under gnotobiotic conditions and documented. So far, no information is available in the literature on the natural and/or induced infection of cereals in general, and in wheat in particular, with G. diazotrophicus.

Materials and methods Basic culture medium H. H. Youssef · M. Fayez · M. Monib · N. Hegazi ()) Environmental Studies and Research Unit (ESRU), Faculty of Agriculture, Cairo University, 12613 Giza, Egypt e-mail: [email protected] Tel.: +20-2-5728483

Sugarcane plants were collected from different ancient agricultural fields of Giza. Plant materials were separated into roots, stems and leaves, then surface-sterilized by different procedures (Table 1). Semisolid N-deficient medium was inoculated with serial dilutions of root and stem samples. The medium used, a modified LGIP medium (Cavalcante and Dobereiner 1988), contained: 100 g

392 Table 1 Most probable number ( MPN) of endophytic Gluconacetobacter in sugarcane root and stem samples sterilized by various procedures

Treatments

References

Root

Shoot 4

MPN (10 cells g-1) Tap water Chloramine T (1.0%, 5 min) H2O2 (3.0%, 5–10 min) Mercuric chloride (0.1%, 5–10 min) Sodium hypochlorite (3.0%, 3–5 min) Sodium hypochlorite (5.0%, 30 min) C2H5OH (95%, 5–10 s)+sodium hypochlorite (3.0%, 30 min) Mercuric chloride (0.2% in 50.0% C2H5OH, 4 min) Sodium hypochlorite (6.0%, 2 min)+C2H5OH (95.0%, 15 s)+flamed

sucrose l-1; 0.2 g K2HPO4 l-1; 0.6 g KH2PO4 l-1; 0.2 g MgSO4.7H2O l-1; 0.02 g CaCl2.2H2O l-1; 0.002 g Na2MoO4.2H2O l-1; 0.01 g FeCl3.6H2O l-1; 0.5% bromothymal blue in 0.2 M KOH, and 2.0 g agar l-1. The pH was adjusted to 5.5. Enumeration of Gluconacetobacter in sugarcane samples Gluconacetobacter were isolated on modified semi-solid N-free LGIP medium and the population densities were determined. Tubes of culture medium were inoculated with aliquots of a series of serial dilutions prepared for homogenates of plant parts in the medium described by Reinhold et al. (1985) lacking C and N. Tubes were considered positive when showing, after 5 days incubation at 30C, the typical dark-orange surface pellicle of Gluconacetobacter and the clear, colourless medium below. Most probable numbers (MPN) were calculated according to Alexander (1982). Representative isolates were obtained by single-colony isolation on agar plates of the previous medium. After 7–10 days, pure orange colonies were transferred onto LGIP medium. One hundred and eighty-six pure isolates were examined for C2H2-reducing activity according to Hegazi et al. (1980) and then screened to give ten potential N2-fixing ones. Selected isolates were identified using methods described by Krieg and Holt (1984). The API microtube systems, API 20 E and API 20 NE, were used as a standardized micromethod (Logan and Berkeley 1984) in addition to conventional tests such as the Gram test and motility. A G. diazotrophicus type culture (ATCC 49037) was used as a reference strain. The isolates obtained, in addition to the type culture strain, were grown in LGIP medium where sucrose was substituted by cane juice at 10% concentration. Cultures were measured for C2H2 reducing activity.

Cavalcante and Dobereiner (1988) Paula et al. (1991) Somasegaran and Hoben (1985) Bilal et al. (1990) Somasegaran and Hoben (1985) Gagne et al. (1987)

12.6

315.0

21.6 820.0 3.0 2.2 24.3 1.0

24.3 1440.0 1.6 15.3 24.3 1.6

1.6

1.6

12.6

0.02

the introduced diazotrophs into the plant root. For the other set of tubes, seedlings were not subjected to initial uprooting. Tubes were incubated in a growth chamber with a day/night cycle of 8/16 h with temperatures of 21/19C, respectively. Seedlings were sampled when 21 days old, surface sterilized and plant roots and stems were analysed for Gluconacetobacter population densities using the MPN technique. To measure C2H2reducing activities, whole plant samples were aseptically transferred to test tubes, sealed with rubber stoppers and 10% of the gas phase was replaced by C2H2. Then, tubes were incubated at 30C for 24 h. The C2H4 produced was measured by injecting 0.5 ml of the gases into a DELSI DN 200/250 gas chromatograph (Hegazi et al. 1980). Plant biomass yield was determined after drying at 70C to constant weights. The presence of Gluconacetobacter as an endophyte was studied by light and scanning electron microscopy. Roots and stems were excised and cross-sections were carefully hand-prepared. They were incubated overnight in tubes half filled with tetrazoliumphosphate buffer solution (PBMT). This solution consisted of 0.05 M potassium phosphate buffer (pH 7.0) containing 0.625 g malate l-1 and 1.5 g 2,3,5-triphenyltetrazolium chloride l-1. The buffer-malate mixture was autoclaved and then tetrazolium added. The PBMT-treated plant segments were placed in water under a cover slip, examined microscopically under a bright field and photographed (Patriquin and Dobereiner 1978). For scanning electron microscopy (Harley and Fergusen 1990), root or stem sections were cut, fixed in 2.5% glutaraldehyde for 24 h at 4C, and then post-fixed in 1% osmium tetroxide for 1 h at room temperature. The specimens were then dehydrated with increasing concentrations of acetone, critical-point dried, and finally sputtercoated with gold. The examination, measurements and photographing were done by using a Jeol scanning electron microscope (JSMT 330 A) equipped with an image recording and processing system (SemAfore).

Endophytic establishment of Gluconacetobacter in wheat plants One of the local isolates (VIS 8) and the type culture strain (ATCC 49037) were tested for their ability to invade the root system of wheat cv. Gemaza 3 in a gnotobiotic culture. Tubes of 2.5 cm diameter and 14.5 cm length filled with 25 ml semi-solid medium (see Reinhold et al. 1985) lacking C and N, and containing fine sand (1 mm diameter, 20 g tube-1) were autoclaved and thoroughly mixed. Healthy seedlings which developed from surface-sterilized seeds on the nutrient agar plates were transferred into tubes at two seedlings per tube. They were top-inoculated with 1-ml portions (108 cells) of either the local VIS 8 isolate or the ATCC strain. Two sets of culture tubes were prepared. For the first set, seedlings were vertically uprooted twice by completely raising the root system then immersing it again into the same culture medium. This procedure probably creates cracks on the root surface through friction with sand particles, which should facilitate the invasion of

Results Ecology of Gluconacetobacter spp. The majority of the surface-sterilization procedures applied, except for the H2O2 treatment, were somewhat similar in eliminating a great portion of root and stem surface-colonizing microorganisms (Table 1). Among them, treatment with 95% ethanol for 5–10 s followed by 3% sodium hypochlorite (30 min) seemed to be the most appropriate and successfully facilitated the gentle isolation of the endophytic Gluconacetobacter diazotrophs from the internal root and stem tissues of plants.

393 Table 2 Physiological properties of representative endophytic local Gluconacetobacter-like isolates obtained from sugar cane using API 20 E. Reading of API galleries extended for 36 h of the incubation. R Root, S stem origin

Table 3 Physiological characteristics of local Gluconacetobacter -like isolates based on their API 20 NE profile. Reading of API galleries extended for 36 h of the incubation. For abbreviations, see Table 2

Characters

b-Galactosidase Arginine dihydrolase Lysine decarboxylase Ornithine decarboxylase Citrate utilization H2S production Urease Tryptophan deaminase Indole production Acetoin production Gelatinase Fermentation of Glucose Mannitol Inositol Sorbitol Rhamnose Sucrose Melibiose Amygdalin Arabinose Cytochrome-oxidase NO2 production N2 production Motility MacConkey Fermentation of glucose Oxidation of glucose

Characters of NO3NO3- to

NO2-

to Reduction Reduction N2 Indole production Glucose (acidification) Arginine dihydrolase, urease Hydrolysis (b-glucosidase) Hydrolysis (protease) b-Galactosidase Glucose Arabinose Mannose Mannitol N -Acetyl-glucosamine Maltose Gluconate Caprate Adipate Malate Citrate Phenylacetate Cytochrome oxidase Dinitrogenase activity (nmoles C2H4 culture-1 h-1)

Gluconacetobacter spp. were encountered in dense populations inside root and stem tissues of sugar cane with population densities ranging from 1.0 to 820.0104 cells g-1 root and from 0.02 to 1,440.0104 cells g-1 stem. This indicates the endophytic nature of the organism in root and stem tissues.

Gluconacetobacter (ATCC 49037)

Local isolates IR

VI R

VI S

IX R

+ +  + +     + 

+    +     + 

+   + +     + +

+ +  + +     + 

+ +  + +     + 

+ +  + + + + + + +   + + + +

+ +  + + + + + +  +  + + + +

+ +  + + +  + +  +  + + + +

+ +  + + + + + +  +  + + + +

+ +  + + + + + +  +  + + + +

Gluconacetobacter (ATCC 49037)

Local isolates IR

VI R

VI S

IX R

+ +  + +   + + + + + + + +   + +   2.56

+  +   +  + + + + + + + +   + +   3

+  +   + € + + + + + + + +   + +   8

+  + +  +  + + + + + + + + € € + + €  20

+  + +  + € + + + + + + + + +  + + +  47

The typical dark-orange colonies developed on LGIP agar plates were purified. The most active isolates for C2H2 reduction (ten isolates) were selected for further identification. In general, cells of tested isolates were small, motile or non-motile, Gram-negative, aerobic rods showing pellicle

394 Fig. 1 Cumulative C2H2-reducing activity of Glucanacetobacter -like isolates and type culture

Table 4 Population densities of endophytic Gluconacetobacter -like diazotrophs in roots, stems and leaves of sugar cane plants of various ages determined by MPN and plate count techniques on LGIP. cfu Colony-forming units, ND not detected Plant age

Total endophytes

Plate count

Gluconacetobacter -like (dark-orange colonies)

Total endophytes

1 Year old 2 Years old 3 Years old Site I Site II Site III

MPN (103 cells g-1)

(103 cfu g-1)

Roots

Stems

Roots

2,000 290

7.0 3.0

2,300 1,170 7,400

30.0 ND 120

Leaves

Stems

0.6 ND

8,600 440

54 27

0.4 0.03 1.1

86,000 5,900 340

71 710 200

Leaves 09 ND 2.0 300 69

formation (micro-aerobic aerotaxis) in semi-solid Ndeficient medium with 10% sucrose. They formed a thick surface pellicle after 7–10 days of incubation at 30C. Gluconacetobacter -like isolates exhibited C2H2-reducing activities in the range of 3.2–59.9 nmoles C2H4 culture-1 h-1. Further identification was performed with the API microtube systems, API 20 E strip (Table 2) for Enterobacteriaceae and API 20 NE profile (Table 3) for nonEnterobacteriaceae. The G. diazotrophicus type culture (ATCC 49037) was found to be related to Enterobacter cloacae and Pseudomonas luteala based on API 20 E and 20 NE respectively. The majority of local Gluconacetobacter -like diazotrophs fitted the same description. This indicates that such isolates are closely related to the type culture G. diazotrophicus. The comparative C2H2-reducing activity of Gluconacetobacter -like isolates and the type culture was measured in LGIP medium containing 10% (v/v) cane

Gluconacetobacter -like (dark-orange colonies)

(%) a

Roots

Leaves

Roots

Stems

Leaves

Stems

390 50

9.0 10.0

1.0 ND

4.53 11.4

16.7 37.0

11.1 0.0

36,000 2,750 80

5.0 280 90.0

0.9 3.0 30.0

41.9 46.6 23.5

7.04 39.4 45.0

45.0 1.00 43.5

juice instead of sucrose. Figure 1 demonstrates that all tested diazotrophs successfully reduced C2H2 and produced appreciable amounts of C2H4 in the presence of cane juice. This presents additional evidence that the local isolates are closely related to G. diazotrophicus. Among all isolates, VIS 8 was found to be an active N2-fixer (59.9 nmoles C2H4 culture-1 h-1) and therefore was used in the inoculation experiment. The occurrence of Gluconacetobacter spp. as an endophyte was further checked in an additional number of sugarcane plants of different ages taken from various sites of the Giza fields. Treatment with C2H5OH followed by sodium hypochlorite (treatment VII, Table 1) was applied for surface sterilization of the plant parts. Population densities of the endophytic diazotrophs determined by MPN and plate count techniques are presented in Table 4. Total endophytes were encountered in high numbers of 102–>107 g-1. Roots hosted exceptional densities compared to stem tissues. The microorganisms

395 Table 5 Average root and stem dry weights and population of Gluconacetobacter of 21-day-old wheat seedlings inoculated with Gluconacetobacter spp. LSD Least significant difference Treatments

Uprooting Without uprooting (A) With uprooting (B) LSD (0.05) LSD (0.01) Inoculation Non-inoculated (C) Acetobacter type culture (D) -Local Acetobacter isolate (E) LSD (0.05) 0.01 Interactions AC BC AD BD AE BE LSD (0.05) LSD (0.01)

Dry weights

Gluconacetobacter population

(mg per six plants)

(log no. per plant)

Root

Shoot

Root

Shoot

41.9 33.1 12.8 17.5

115.9 100.3 12.49 17.04

1.46 2.61 0.22 0.32

1.75 2.35 0.56 0.79

53.5 27.6

136.4 78.4

0.00 3.59

0.00 4.16

31.5

109.6

2.52

1.99

0.18 0.26

0.68 0.97

0.00 0.00 2.69 4.48 1.69 3.34 0.32 0.45

0.00 0.00 3.45 4.86 1.79 2.20 0.96 1.37

10.5 14.3 53.5 53.5 36.0 19.2 36.2 26.7 18.5 24.8

10.19 13.91 136.4 136.4 79.1 77.7 132.3 87.0 17.7 24.1

were rare in sugarcane leaves, present at densities of <103 g-1, or completely absent. In the majority of cases and regardless of the sugarcane organ examined, Gluconacetobacter -like diazotrophs developing orange colonies on LGIP agar plates represented up to 46.6% of the total endophytes. There was a tendency towards an increasing population density of endophytes with increasing of plant age. Verifying the endophytic nature of Gluconacetobacter spp. isolates Cells of Gluconacetobacter spp. were introduced into seedlings of wheat grown in gnotobiotic cultures. In general, inoculation did not alter the biomass production of wheat plants and even their development was at the expense of plant growth (Table 5). Initial shaking of the root system with sand particles decreased the dry weights of roots and stems. Inoculated Gluconacetobacter did establish in the gnotobiotic system (Table 5). The initial uprooting of seedlings at the time of planting resulted in a significant colonization of roots and shoots by the introduced Gluconacetobacter. As for interactions, the highest and statistically significant endophytic populations were recorded for initially uprooted seedlings inoculated with the type culture, followed by the local isolate. Examinations with both optical and scanning electron microscopy showed that cells were endophytic in all samples tested.

Fig. 2 Scanning electron micrograph of a 21-day-old wheat stem inoculated with Gluconacetobacter diazotrophicus without ( A) and with ( B) initial shaking with sand. A Cross-section illustrating the xylem vessels filled with slime cells, B slime layers of bacteria filling xylem parenchyma

This was observed in apparently intact, enlarged epidermal cells of the root and also in stem tissues, indicating the ability of the bacterium to migrate upward in the transpiration stream to the shoot tissues. Figures 2, 3 and 4 show infection of the interior of parenchyma cells of the cortex and vascular cylinder of wheat roots and stems. Initial uprooting facilitated an intense colonization of inner plant tissues.

Discussion G. diazotrophicus, originally isolated by Cavalcante and Dobereiner (1988), has been further reported in sugarcane plants grown in Mexico, Cuba and Australia (Li and MacRae 1992). Population densities ranged from 103 to 107 colony-forming units g-1 fresh weight. They were traced in roots, basal and apical stems, leaves and in plant debris (Dobereiner et al. 1988). This particular diazotroph

396 Fig. 3A–D Scanning electron micrographs of 21-days-old wheat roots, inoculated with the local isolate VIS8 and initially uprooted. A General aspect of the whole section; B bacterial cells and mucus-filled parenchyma in the cortex and vascular cylinder; C, D higher magnification illustrating the intense invasion of bacterial cells into the parenchyma cells of the vascular cylinder

has a number of interesting properties, some of which make it a likely candidate for providing fixed N2 to sugarcane. These characteristics include: (1) the ability to grow on 10% sucrose as the sole C source, (2) a very low pH (5.5) requirement for optimum growth, (3) no or little inhibition of NO3- and NH4+ of the bacterial nitrogenase activity, and (4) a capability for microaerobic N fixation (Stephan et al. 1991; Burris 1994). These are a group of properties that distinguish G. diazotrophicus from Azospirillum spp. The present paper presents previously unpublished information on Gluconacetobacter-like bacteria associated with sugarcane traditionally cultivated in old Egyptian fields. Among the tested methods of surface sterilization for sugarcane parts, the combined treatment with C2H5OH and sodium hypochlorite was efficient for the elimination of contaminants and hence facilitated the isolation of the diazotroph. Based on MPN estimates, the diazotroph was found in all tested samples with as many as 105–107 cells g-1 root or stem, indicating the internal colonization of sugarcane by the bacterium. Similar high numbers were encountered in sugarcane grown in Australia as assessed with an indirect enzyme-linked immunosorbent assay (Li and MacRae 1992). The endophytic nature of G. diazotrophicus, reported previously by Dobereiner et al. (1988) was confirmed, as this diazotroph was present in roots, stems and aerial parts collected from various cultivars at different stages of growth and maturation. Population densities in all plant parts ranged from 103 to 106 cells g-1 fresh weight. The bacterium was also isolated from xylem sap, indicating translocation of the microorganism through the plant tissues in the xylem (Reis 1991).

An improved isolation method described by Reis et al. (1994) was used to examine various samples representing six different forage grasses, sorghum, maize and 11 weed species collected within sugar cane fields. All were negative for G. diazotrophicus. Therefore, an endophytic nature and high specificity were proposed for this bacterium on the basis of the observations showing its exclusive occurrence in sugarcane, sweet potatoes and Cameroon grass. All these plants are propagated vegetatively, and contain high sugar concentrations (Dobereiner et al. 1988). Results of the present study and those of Dobereiner et al. (1988), Dobereiner (1992), Reis et al. (1994), Paula et al. (1991) and Li and MacRae (1992), who reported the occurrence of G. diazotrophicus in roots, tubers and stems of sweet potato, confirm the endophytic nature of G. diazotrophicus in sucrose-rich plants. In an attempt to test the ability of Gluconacetobacter to endophytically colonize other non-legume plants, the reported gnotobiotic culture experiment was designed where wheat was inoculated with the diazotroph. Although Gluconacetobacter inoculation did not exert beneficial effects on the growth of the cereal plant, the results obtained are of particular interest because the tested isolate was capable of colonizing the internal tissues of a non-sugar plant not considered a natural host until now. Similar to the colonization in sugarcane, the invading Gluconacetobacter cells were translocated towards inner ground tissues, xylum and to various parts in the wheat plants.

397

Fig. 4A. B Scanning electron micrographs of 21-day-old wheat roots, inoculated with the G. diazotrophicus type culture together with initial uprooting. A Bacterial cells and their mucus inside xylem vessels, B a magnified portion of A Acknowledgements Thanks are due to Prof. S. El-Deeb at the Plant Pathology Department, the Faculty of Agriculture, Ain Shams University for his unlimited support during the electron microscopic examinations of plant samples. The very important botanical information of Prof. K. El-Sahaar of the Faculty of Agriculture, Cairo University is very much appreciated. This study was supported by the research grant BLAFE /FC/31/3–94 offered by the National Project on Agro-Technologies Based on Biological Nitrogen Fixation for the Development of Sinai Agriculture.

References Alexander M (1982) Most probable number for microbial populations. In: Page AL (ed) Methods of soil analysis, part 2. American Society of Agronomy, Soil Science Society of America, Madison, Wis., pp 815–820 Bilal R, Rasul G, Qureshi JA, Malik KA (1990) Characterization of Azospirillum and related diazotrophs associated with roots of plants growing in saline soils. World J Microbiol Biotechnol 6:46–52

Burris RH (1994) Comparative study of the response of Azotobacter vinelandii and Acetobacter diazotrophicus to changes in pH. Protoplasma 183:62–66 Cavalcante VA, Dobereiner J (1988) A new acid-tolerant nitrogenfixing bacterium associated with sugar cane. Plant Soil 108:23– 31 Cojho EH, Reis VM, Schenberg AC, Dobereiner J (1993) Interactions of Acetobacter diazotrophicus with an amylolytic yeast in nitrogen-free batch culture. FEMS Microbiol Lett 106:341– 346 Dobereiner J (1992) History and new perspectives of diazotrophs in association with non-leguminous plants. Symbiosis 13:1–13 Dobereiner J, Reis V, Lazarini AC (1988) New N2-fixing bacteria in association with cereals and sugarcane. In: Bothe H, de Bruijn FJ, Newton WE (eds) Nitrogen fixation: one hundred years after. Fischer, Stuttgart, pp 717–722 Fuentes-Ramirez LE, Jimenez-Salgado T, Abarca-Ocampo IR, Caballero-Mellado J (1993) Acetobacter diazotrophicus, an indole acetic acid producing bacterium isolated from sugarcane cultivars of Mexico. Plant Soil 154:145–150 Gagne S, Richard C, Rousseau H, Antoun H (1987) Xylem-residing bacteria in alfalfa roots. Can J Microbiol 33:996–1000 Gillis M, Kersters K, Hoste B, Janssens D, Kroppensted’r RM, Stephan MP, Teixeira KRS, Dobereiner J, Deley J (1989) Acetobacter diazotrophicus sp. nov., a nitrogen-fixing acetic acid bacterium associated with sugarcane. Int J Syst Bacteriol 39:361–364 Harley MM, Fergusen IK (1990) The role of the SEM in pollen morphology and plant systematic. In: Claugher D (ed) Scanning electron microscope in taxonomy and functional morphology. (Systemics Association special volume) Clarendon press, Oxford, pp 45–68 Hegazi NA, Amer HA, Monib M (1980) Studies on N2-fixing spirilla ( Azospirillum spp.) in Egyptian soil. Rev Ecol Biol Soil 17:491–499 Krieg NR, Holt JG (1984) Bergey’s manual of systematic bacteriology, 1st edn. Williams and Wiilkins, Baltimore, Md. Li RP, MacRae IC (1991) Specific association of diazotrophic acetobacters with sugarcane. Soil Biol Biochem 23:999–1002 Li RP, MacRae IC (1992) Specific identification and enumeration of Acetobacter diazotrophicus in sugarcane. Soil Biol Biochem 24:413–419 Logan NA, Berkeley RCW (1984) Identification of Bacillus strains using the API system. J Can Microbiol 130:1871–1882 Patriquin DG, Dobereiner J (1978) Light microscopy observations of tetrazolium-reducing bacteria in the endorhizosphere of maize and other grasses in Brazil. Can J Microbiol 24:734–742 Paula MA, Reis VM, Dobereiner J (1991) Interactions of Glomus clarum with Acetobacter diazotrophicus in infection of sweet potato ( Ipomoea batatas), sugar cane ( Saccharum spp.), and sweet sorghum ( Sorghum vulgare). Biol Fertil Soils 11:111– 115 Reinhold B, Hurek T, Fendrik I (1985) Strain-specific chemotaxis of Azospirillum spp. J Bacteriol 162:190–195 Reis VM (1991) MSc thesis. UFRRJ, Seropedica-RJ, Brazil, 119 p Reis VM, Olivares FL, Dobereiner J (1994) Improved methodology for isolation of Acetobacter diazotrophicus and confirmation of its endophytic habitat. World J Microbiol Biotechnol 10:101– 104 Somasegaran P, Hoben HJ (1985) Methods in legume- Rhizobium technology. In: University of Hawaii NifTAL Project (ed). MIRCEN Department of Agronomy and Soil Science, Hawaii Institute of Tropical Agriculture and Human Resources, College of Tropical Agriculture and Human Resources, Hawaii, pp 7–9 Stephan MP, Oliveira M, Teixeira KRS, Martinez-Drets G, Dobereiner J (1991) Physiology and dinitrogen fixation of Acetobacter diazotrophicus. FEMS Microbiol Lett 77:67–72