Plant and Soil (2005) 270: 223–232 © Springer 2005

DOI 10.1007/s11104-004-1522-7

Isolation and characterization of Azotobacter and Azospirillum strains

from the sugarcane rhizosphere

N. Tejera1 , C. Lluch1 , M.V. Martı́nez-Toledo2 & J. González-López2,∗

1 Departamento de Fisiologı́a Vegetal, Facultad de Ciencias, Universidad de Granada, Campus de Fuentenueva
s/n, 18071 Granada, Spain. 2 Instituto del Agua, Universidad de Granada, 18071 Granada, Spain. 3 Corresponding
author∗

Received 22 January 2004. Accepted in revised form 20 July 2004

Key words: Azospirillum, Azotobacter, nitrogen fixation, sugarcane

Abstract
Bacteria with the ability to grow on nitrogen-free media and with nitrogenase activity under aerobic or microaerobic
conditions were isolated from sugarcane roots collected from four different agricultural locations in Granada
(Spain). Isolates were Gram negative rods and were identified as Azotobacter chroococcum and Azospirillum
brasilense. Our results suggest that Azotobacter isolates do not have a particular affinity for sugarcane rhizospheres
and that, on the contrary, Azospirillum isolates show specific association and perhaps endophytic colonization of
sugarcane. However, obligate endophytes (Gluconacetobacter diazotrophicus) were not found in the apoplastic
fluid of the stems and macerates extracts of sugarcane tissues with the procedure applied. Population of this
microorganism might be in low number in the Spanish sugarcane varieties studied which is also discussed.

Introduction sues (Cavalcante and Döbereiner, 1988; Olivares et al.,

Grasses constituted an important source of food all
over the world. In particular, grasses such as rice,
wheat, maize, sorghum, and sugarcane, currently have
much of their N needs supplied by costly mineral
fertilizers (Döbereiner et al., 1995; Triplett, 1996).
Evidence for biological nitrogen fixation in sugarcane
(Saccharum spp.) was reported in Brazilian sugarcane
varieties. Studies on long-term N-balance and 15 N iso-
tope dilution technique (Urquiaga et al., 1992) have
shown that some sugarcane varieties may actually ob-
tain up to 70% of their N requirements by nitrogen
fixation. In this process seems to participate both rhi-
zosphere and endophytic diazotrophs (Baldani et al.,
1997). N2 fixing bacteria such as Enterobacter cloa-
cae, Erwinia herbicola, Klebsiella pneumoniae, Azo-
tobacter vinelandi, Paenibacillus polymixa, Azospiril-
lum spp., Herbaspirillum spp. and Gluconacetobacter
diazotrophicus colonise the sugarcane plant and its tis-
∗ FAX No: 34-958-243094.
E-mail: jgl@ugr.es
1996).
Azotobacter is able to fix at least 10 mg N per gram
of carbohydrate (Becking, 1992). This bacterium is an
obligate aerobic, although it can grow under low pO2 .
The ecological distribution of Azotobacter spp. is a
complicated subject and is related with diverse fac-
tors which determine the presence or absence of this
bacterium in an specific soil. It has been shown that
the soil characteristics and climate conditions affect
the distribution of this microorganism (Döbereiner and
Pedrosa, 1987; González-López, 1992); it includes or-
ganic matter content, moisture, C/N relation and pH
(González-López, 1992).
Azospirillum species belong to the facultative en-
dophytic diazotrophs group which colonize the surface
and the interior of roots, being this kind of associa-
tion considered as the starting point of most ongoing
BNF programs with non-legume plants worldwide
(Baldani et al., 1997). Bacteria are microaerophilic,
nitrogen-fixing, Gram-negative rods and often as-
sociated with roots of cereals and grasses (Grifoni
et al., 1995). However obligate endophytes such as
224
Gluconacetobacter diazotrophicus and Herbaspiril-
lum spp. seem to be the promissory group in rela-
tion to nitrogen fixation associated with sugarcane.
These bacteria have an advantage over root-associated
diazotrophs, as Beijerinckia spp. and Azotobacter
spp., they colonise the interior rather than the sur-
face of plants, and hence have better possibilities
to exploit carbon substrates supplied by the plant
(Boddey et al., 1995; Sprent and James, 1995;
Triplett, 1996). However, some characteristics distin-
guish G. diazotrophicus as the high sucrose (10%)
tolerance, growth and nitrogen fixation at low pH
(5.0 or less) and that nitrogen fixation is only par-
tially inhibited by NH+ 4 , especially at high sucrose
concentrations (Boddey et al., 1991; Stephan et al.,
1991).
It has been a general practice to apply 250 kg
N ha−1 yr−1 or more in most sugarcane cultivating
countries. In sugarcane fields in Spain farmers fre-
quently use high amounts of fertilizers (400–500 kg
N ha−1 yr−1 ) which are applied 30 days after sprout-
ing, together with other three applications distributed
throughout the crop cycle until harvest. Interactions
and association between microorganism and their host
plants may be inhibited by high levels of added fer-
tilizer (Oak, 1992). Kirchhof et al. (1997) observed
a higher number of diazotrophic bacteria associated
with graminaceous plants if no fertilizer was applied.
In this context Yoneyama et al. (1997) suggested
that N-fertilisation in the early growth stage may not
have a depressing effect on N2 -fixation. The isola-
tion frequency of G. diazotrophicus from sugarcane
plants diminished proportionally with the quantity of
N-fertilisation in the fields (Caballero-Mellado et al.,
1995; Fuentes-Ramirez et al., 1993, 1999; Muthuku-
marasamy et al., 1999, 2002). However the population
of other diazotrophs in sugarcane rhizosphere, par-
ticularly Herbaspirillum spp., was not affected by
the inorganic N-application (Muthukumarasamy et al.,
1999).
In the present work we report the characteriza-
tion of Azotobacter and Azospirillum isolates from
roots of sugarcane collected from different agricul-
tural soils located in the south of Spain. We also
suggest that Gluconacetobacter diazotrophicus popu-
lations associated with this grass should be low in soil,
rhizosphere and sugarcane, thus the contribution of
this microorganism to nitrogen nutrition of this crop,
at least under our agricultural conditions, should not
be significant.
Material and methods
Soil and plant samples
Four sugarcane fields (A, B, C and D) located in
Motril, near Granada (Spain) were sampled. Fields
have been in monoculture for more than ten years and
sugarcane plants were fertilized with different applica-
tions raising 400–500 kg N ha−1 yr−1 . Two kg of soil
samples, separated 50 cm from sugarcane plants, were
collected from the upper 15 cm layer of soil. Also sam-
ples from the rhizosphere soil, root, stems and blades
(varieties NCo310, RD75-11 and TUC77-42) of three
months old plants (ratton sugarcane) were collected,
according to Muthukumarasamy et al. (1999), for the
isolation of nitrogen fixing endophytes. Plants grew
in a typical Mediterranean environment and samples
were taken during May which was a quite irregu-
lar month for climatic characteristics. Maximum and
minimum temperature were 25 and 15 ◦ C, respec-
tively, average air humidity was 75% and precipitation
42 l m−2 .
Soils chemical and granulometric analyses were
performed as described by Soil Conservation Ser-
vice (1975). Nitrate and nitrite concentration in the
soil samples was quantified in water extracts (at 210
and 540 nm, respectively) according to Linblad and
Guerrero (1993), to estimate the remaining of these
inorganic nitrogen forms in soil after fertilization (Ta-
ble 1).
Isolation and culture of nitrogen-fixing bacteria
Burk’s N-free media (Martínez-Toledo et al., 1985)
were used for the isolation of aerobic nitrogen-fixing
bacteria. Soil samples (2 kg) were obtained from agri-
cultural soils without sugarcane plants, and then 2 g
soil samples were added into 500 mL Erlenmeyer
flasks containing 18 mL of Burk’s liquid medium.
Samples were incubated 48 h at 28 ◦ C. Each culture
was subcultured five times after 48 h incubation and
enriched cultures were plated on Burk’s solid medium
and then incubated (28 ◦ C for 48 h). Approximately 10
randomly selected bacterial colonies were subcultured
from each Burk’s agar plate. Single colonies were
restreaked on Burk’s agar plate for further purification.
Burk’s semi-solid media was used for the isolation
of aerobic nitrogen-fixing bacteria from rhizosphere
of sugarcane. Sections of fresh roots were placed
on the agar media and incubated at 28 ◦ C for 72 h
 225
Table 1. Mechanical and chemical properties of soils cultured with sugarcane
varieties (NCo310, RD75-11 and TUC77-42) from Spain.

Soil A B C D

Type Typical Calcixerollic Mollic Typical

xerorthent xerochrept xerofluvent xerorthent

Texture Loam Silty loam Silty loam Silty loam

Sand 73.7 33.3 34.4 44.8
Silt (%) 19.8 52.4 54.1 43.8
Clay (%) 6.5 15.3 12.5 12.6
pH (H2 O) 8.67 7.82 8.41 8.15
Organic matter (%) 1.19 2.72 2.42 2.02
Total N (%) 0.07 0.19 0.14 0.15
Nitrate 2.06 25.12 4.99 24.51
(mg kg−1 )
Nitrite 0.020 0.029 0.032 0.046
(mg kg−1 )

Values are average of three separate determinations.

according to Martínez-Toledo et al. (1985). Bacteria
growing out from the root were transferred to plates of
the same medium for their purification. Ten randomly
selected bacterial colonies were subcultured from each
Burk’s solid plate.
The isolation of microaerobic nitrogen-fixing bac-
teria from soil, rhizosphere and root samples were
carried out according to Rodríguez-Cáceres (1982).
Samples of each type (0.5 g) were placed in semi-solid
medium (nitrogen-free malate medium) containing
0.175% (w/v) agar (Gómez et al., 1998) and then incu-
bated at 30 ◦ C for 48 h under microaerobic conditions
(Haahtela et al., 1981). Each culture was subcul-
tured five times after 48 h incubation and enriched
cultures were plated on NH+ 4 -malate solid medium
amended with 2% (w/v) agar and incubated for 48 h
at 30 ◦ C under aerobic conditions. Selected bacter-
ial colonies (approximately 10) were subcultured from
each NH+ 4 -malate solid plate.
For the isolation of Gluconacetobacter diazotroph-
icus, tissues samples were disinfected with 70% alco-
hol for 30 s followed by washing with sterile water,
and again surface-sterilised with 1% chloramine T for
3 min (Muthukumarasami et al., 1999). One gram of
root samples and 10 g in the case of stems and leaves
were homogenised in 100 mL sterile water and used
for serial dilutions. Also apoplastic sap, extracted by
centrifugation at 3,000 g for 20 min of the stems pieces
previously dipped in ethanol and surface flamed, was
used as source of inoculation (Dong et al., 1994). Bac-
teria were isolated using 5 mL semi-solid medium
in 10 mL vials containing LGIP medium (pH 5.5)
with 10% of cane sugar as carbon source and 5 mL
L−1 of cane juice (Reis et al., 1994). Vials with the
semi-solid LGI-P media were inoculated with appro-
priate serial dilutions of 0.1 mL of root, stem and
leaf extracts according to Muthukumarasami et al.
(1999) or 0.1 mL of stems apoplastic fluid (Dong
et al., 1994) and were kept at 30 ◦ C for 5–7 days.
Further, yellowish pellicles formed in the semi-solid
medium were streaked on LGIP solid plates (with 10%
sugar) supplemented with 0.1% yeast extract and 2.5%
agar. Colonies, which were initially white and later
became yellowish; were then transferred to a semi-
solid medium and further streaked on the purification
medium supplemented with 0.2% NH4 Cl and 0.2%
yeast extract.
Characterization of aerobic isolates
Pure isolates of the aerobic nitrogen-fixing bacteria
from soil, rhizosphere or root were characterized us-
ing the criteria of Bergey’s Manual of Systematic
Bacteriology (1984) and Bergey’s Manual of Determi-
native Bacteriology (1994). The following morpholog-
ical, physiological and biochemical tests were used:
Colony morphology, size, Gram staining, production
of diffusible and non-diffusible pigments were deter-
mined on Burk’s solid medium after 2 and 5 days
of incubation at 30 ◦ C. Motility was determined in
wet mounts and flagella arrangement assessed by the
technique of Rhodes (1958).
226
Encystment was induced by the method of So-
colofsky and Wyss (1961). The cyst were stained
by the method of Vela and Wyss (1964). Poly-β-
hydroxybutyric acid (PHB) granules were examined
according to the method described by Baker (1967).
Utilization of glucose, rhamnose, caproate, capry-
late, meso-inositol, mannitol and malonate as carbon
source was assayed on Burk’s basal medium with a
final concentration of 0.5% (w/v) of each substance.
Starch hydrolysis was tested in cultures on Burk’s
solid medium containing 1% (w/v) potato starch by
flooding with Lugol’s iodine. Oxidase production
and catalase activity were examined on Burk’s solid
medium after 48 h incubation at 30 o C.
Growth at different initial pH values was assessed
by absorbance measurements (540 nm) after 48 h
incubation on liquid Burk’s medium with the pH ad-
justed to 4.0, 5.0, 6.0, 7.0, 8.0 or 9.0 with HCl or
KOH.
Reference strains of Azotobacter chroococcum
CECT 4435 and A. vinelandii ATCC 12837 as controls
were included for comparison.
Characterization of microaerobic isolates
Pure isolates of the microaerobic nitrogen-fixing bac-
teria from soil, rhizosphere or root were characterized
according to Bergey’s Manual of Systematic Bacteri-
ology (1984) and Bergey’s Manual of Determinative
Bacteriology (1994). The following tests were per-
formed: Gram staining, motility and microaerobic
growth were examined in semi-solid N-free malate
medium. Flagellar arrangement on MPSS agar at
30 ◦ C was observed by the technique of Rhodes
(1958). Production of pigments was assessed on
BMS agar after 2 weeks at 35 ◦ C (Krieg and Holt,
1984).
Intracellular accumulation of PHB on semi-solid
N-free malate medium was examined according to
Baker (1967) after 1 week of incubation at 30 ◦ C. Cyst
formation was detected according to Vela and Wyss
(1964).
Oxidase and catalase activities were examined on
cultures grown on malate agar medium after 72 h
at 30 ◦ C. Utilization of glucose, α-ketoglutarate and
mannitol as sole carbon sources was detected in N-free
malate media. Starch hydrolysis was tested in cultures
on malate solid medium containing 1% (w/v) potato
starch, by flooding with Luggol’s iodine.
A reference strain of Azospirillum brasilense
ATCC 21145 was included for comparison.
Characterization of isolates from apoplastic sap and
macerates of sugarcane tissues
Pure isolates from the apoplastic fluid of the stem and
macerates of sugarcane tissues were identified based
on specific physiological, biochemical and molecular
tests for G. diazotrophicus. Oxidation of ethanol and
glucose and fermentation of various sugars were de-
termined according to Hayward (1964). Bromothymol
blue assimilation in LGIP medium (Reis et al., 1994),
dark brown colonies formation in potato P medium
and growth and N fixation with 30% glucose or su-
crose at a pH lower than 5.5 were assayed (Stephan
et al., 1991). Besides utilization and acidification of
different C substrates (gluconate, methanol, starch,
mannitol, malic and aspartic acid) were also tested
according to Li and McRae (1991) and Ureta et al.
(1995). Reference strain of G. diazotrophicus PAL-5
was included for comparison.
The molecular test of amplifying DNA and analy-
sis by PCR was used to detect the presence of the
bacterium in isolates and tissues of sugarcane. Spe-
cific primers 1440 (5 -GTTGGCTTAGAAGCAGCC-
3 ) and AD (5 -TGCGGCAAAAGCCGGAT-3 ) gen-
erated a 411-bp product following the procedure and
temperature program profile described by Kirchhof
et al. (1998). Extracts from sugarcane plants (30 d
old) cultured in a controlled environmental chamber
and inoculated with G. diazotrophicus PAL-5 strain
were used as a positive control for PCR amplifica-
tion. In addition cultures of PAL-5 strain were also
subjected to PCR reaction. All amplification products
were separated electrophoretically in 1% agarose gels.
Nitrogenase activity
The nitrogenase activity (acetylene reduction) assay
was carried out in 10 mL vials containing 5 mL of
the medium (see below). Isolates from Burk’s N-free
media were grown in this same medium, and isolates
from NH+ 4 -malate and LGIP media were grown in
semi-solid medium N-free malate and LGIP respec-
tively for 48 h at 30 ◦ C. Each vial was sealed with
rubber stopper and the head space (5 mL) was injected
with 10% (v/v) acetylene (Muthukumarasami et al.,
1999). Gas samples (0.2 mL) were removed after 1 h
and assayed for ethylene production with a gas chro-
matograph (Konic model). The chromatograph was
fitted with a 200 cm Poropak-R column and a hydro-
gen flame ionisation detector. Acetylene was gener-
ated immediately before use from calcium carbide and
water. Ethylene contamination of the acetylene was

always known and accounted for final calculations.
Values were expressed as nmoles C2 H4 h−1 culture−1 .
Rep- PCR and ARDRA analyses
For repetitive sequence analysis (Rep-PCR) bacterial
cells were suspended in 20 µL lysis buffer (0.05 M
NaOH, 0.25% SDS) and boiled for 15 min according
to Herrera-Cervera et al. (1999). The lysate was di-
luted in 200 µL distilled sterile water and spun in a
microcentrifuge for 5 min. 2 µL of the supernatant was
used directly for PCR amplification using 50 pmol of
each primer (REP1R-I 5 -IIIIICGICGICATCIGGC-
3 , RE2-I 5 -ICGICTTATCIG-GCCTAC-3 ), 1.25 mM
dNTPs and 2 U Taq-polymerase in a total reaction
volume of 25 µL (de Bruijn, 1992). The following
temperature profile was used: one cycle (6 min at
95 ◦ C), 30 cycles (1 min at 94 ◦ C, 1 min at 40 ◦ C
and 8 min at 65 ◦ C) and one cycle (6 min at 65 ◦ C).
Restriction fragment analysis of PCR-amplified
16S rDNA (ARDRA) was performed basically as de-
scribed previously by Laguerre et al. (1994). The fol-
lowing enzymes were used: TaqI, NdeII and MspI. Re-
striction patterns were compared with reference strains
of Azotobacter chroococcum CECT 4435 and Azospir-
illum brasilense ATCC 21145, obtained from Spanish
collection. The lysate (prepared as described above)
was used directly for PCR amplification using 10 pmol
of each the primer (see below), 200 µM dNTPs and
2 U Taq polymerase in a total reaction volume of
50 µL. PCR amplification involved primers 41f (5 -
GCTCAGATTGAACGCTGGCG-3 ) and 1488r (5 -
CGGTTACCTTGTT-ACGACTTCACC-3 ) for Azoto-
bacter spp. isolates (Herrera-Cervera et al., 1999), and
fD1 (5 -AGAGTTTGATCCTGGCTCAG-3 ) and rD1
(5 -AAGGAGGTGATCCAGCC-3 ) for Azospirillum
spp. isolates (Vinuesa et al., 1998). Amplifications
were carried out with the following temperature pro-
file: 2 min at 94 ◦ C, 10 cycles of denaturalization (40 s
at 94 ◦ C), annealing (1 min at 60 ◦ C, lowering 1 ◦ C
cycle) and extension (2 min at 72 ◦ C), 25 cycles of de-
naturation (40 s at 94 ◦ C), annealing (1 min at 50 ◦ C)
and extension (2 min at 72 ◦ C), and a final extension
for 3 min at 72 ◦ C. Reaction products of both Rep-
PCR and ARDRA analyses were separated in agarose
gels electrophoresis at 1.5 and 3% w/v, respectively.
Results and discussion
Several types of microorganisms were isolated from
soil in four sugarcane fields and rhizosphere samples
227
on N-free medium under aerobic conditions. These
bacteria were transferred successively to solid (N-free)
media and purified. In these experiments, more than
1375 isolates (60% from soil samples and 40% from
rhizosphere samples) with rapid growth (24 h) on
Burk’s N-free media at 28 o C were isolated from four
sugarcane fields. The number of isolates was similar
in soils B, C and D (data not shown). However a lower
number of isolates with rapid growth on Burk’s N-free
media was observed in soils and rhizosphere samples
obtained from soil A. Soil A showed the highest pH
value (8.7) and the lowest organic matter content (Ta-
ble 1); González-López (1992) has suggested that the
optimum pH required for the growth of Azotobacter
spp. is near to 7. Randomly selected bacterial colonies
(137) were subcultured from the 1375 isolates for
further taxonomic studies.
The colonies formed by these bacteria on N-free
media were slightly viscous, semi-transparent during
the early growth and later dark brown. Bacteria were
Gram-negative with rounded ends, with an average
cell size of 1.3 × 4.0 µm after 24 h growth in N-
free liquid culture. On the agar medium they formed
coccoid cells, even at a very early stage. Biochemical
and morphological characteristics of these bacteria in-
cluded the following: motile with peritrichous flagella;
PHB granules positive and starch hydrolysis positive.
Utilization of mannitol, caproate, caprylate, meso-
inositol, malonate but not rhamnose was detected.
Cysts were produced in Burk’s agar medium con-
taining 0.3% n-butanol. Bacteria grew well in Burk’s
liquid medium at pH 9.0, 8.0 and 7.0, but failed to
grow at pH below 6.0.
Isolates (60% from soil samples and 40% from
rhizosphere samples) were classified according to
Bergey’s Manual of Determinative Bacteriology
(1994) as Azotobacter chroococcum and their nitroge-
nase activities were determined by checking the acety-
lene reduction activity of five isolated bacteria from
rhizosphere and soil samples (Figure 1). Amounts of
acetylene reduced by A. chroococcum isolates from
soil or rhizosphere samples were quite different. Rates
obtained in isolates were in the range of 79.6 to
329.5 nmol C2 H4 h−1 culture−1 .
For Azotobacter isolates, there were more isolated
from soil (60%) than from rhizospheres (40%), in-
dicating that there was no particular affinity of them
for sugarcane rhizospheres. In addition, Azotobacter
isolates were not found in the inside of the sugar-
cane roots independently of the field studied (Table 1).
This is not surprising because Azotobacter species
228
Figure 1. Acetylene reduction activity of A. chroococcum CECT
4435 (reference strain), isolates from soil (AS1 , AS2 , AS5 , BS4
and CS1 ) and rhizosphere (AR2 , AR3 , BR2 , BR3 and BR4 ) of
sugarcane. Values are means of five replicates.
are mainly free-living and found ubiquitously in soils
throughout the world.
More than 250 different isolates able to grow on
N-free malate medium at 30 ◦ C under microaerobic
conditions were isolated and purified from rhizosphere
soils and roots of sugarcane. Randomly selected bac-
terial colonies (25) were subcultured from the 250
isolates for further taxonomic studies.
The colonies of these bacteria on malate medium
were opaque and non slimy. Bacteria were Gram-
negative, with a lateral and polar flagella, motile,
slightly-curved and straight rods, about 1.0 × 3.8 µm
of cell size. Biochemical and morphological charac-
teristics of these bacteria included the following: PHB
granules positive, oxidase positive, catalase positive,
cysts formation negative, esculin hydrolysis and phos-
phatase positive, pink pigment production on BMS
agar, starch hydrolysis negative, gluconate positive
and not glucose, α-ketoglutarate and mannitol utiliza-
tion. Biotin was not required for growth.
These isolates (15 from rhizosphere soils and 10
from roots) were classified according to Bergey’s
Manual of Determinative Bacteriology (1994) as
Azospirillum spp. and they showed nitrogenase ac-
tivity (acetylene reduction activity under microaero-
bic conditions). Maximum nitrogenase activities from
four selected bacteria (Table 2) were in the range of
91.5 to 135.1 nmol C2 H4 h−1 culture−1 .
The finding of Azospirillum isolates from the rhi-
zosphere and from surface disinfected roots of sug-
arcane does suggest that this organism colonizes
the surface and interior of roots of sugarcane from
Table 2. Maximum nitrogenase activity (ethylene production)
by selected Azospirillum brasilense isolates from rhizosphere of
sugar cane.
Isolates Field Nitrogenase activity
(nmol C2 H4 h−1 culture−1 )
ATCC 21145 Reference strain 149.6 ± 3.7
AR8 A 91.5 ± 6.4
BR1 B 105.2 ± 5.2
CR3 C 135.1 ± 7.6
DR1 D 111.4 ± 3.6
The values are means of five replicates followed by the standard
deviation.
agricultural soils in Spain. Organisms were not de-
tected in soil samples without roots, suggesting a spe-
cific association and perhaps endophytic colonization
of sugarcane by these organisms.
The identification of both A. chroococcum and
Azospirillum spp. isolates was completed by molecu-
lar tests of repetitive extragenic palindromic sequences
(Rep-PCR) and by restriction analysis of rDNA am-
plified genes (ARDRA). The analysis of patterns gen-
erated by Rep-PCR showed common bands (PCR
products of analogous mobility) from A. chroococcum
isolates (Figure 2, lanes 1–7) and minor differences
with the reference strain CECT 4435. In addition,
Azospirillum spp. isolates showed a similar Rep-PCR
pattern to that detected in the reference strain ATCC
21145 and minor divergent bands were detected in
AR8 and BR1 isolates. The method used led to a rapid
identification of the Azotobacter and Azospirillum iso-
lates, and as was previously reported by de Bruijn
(1992) and Vinuesa et al. (1998) is highly specific for
each isolates and useful for classification of bacterial
strains.
16S rDNA was amplified from six A. chroococcum
isolates from soil samples (AS1 , AS2 and BS4 isolates)
and rhizosphere isolates (AR2 , BR3 and BR4 isolates),
and from four Azospirillum spp. isolates from rhi-
zosphere soils (AR8 ) and roots (BR1 , CR3 , DR1 ). In
all cases isolates produced ARDRA patterns typical
of their A. chroococcum and A. brasilense when they
were compared with the patterns of reference strains
(Figure 3). According to these results we confirm the
presence of A. chroococcum and A. brasilense species
in isolates. Thus, the restriction with the enzymes
Nde II and Taq I produce a reproducible ARDRA pat-
tern for A. chroococcum and A. brasilense respectively
which represent a simple, rapid and reliable tool for
identification of these species. Related to this, DNA

Figure 2. Rep-PCR fingerprint patterns of five A. chroococcum and four A. brasilense isolates from soil and rhizosphere samples of sugarcane.
Lane 1 to 7 show PCR products A. chroococcum isolates CECT 4435 (reference strain), AS1 , AS2 , BS4 , AR2 , BR3 , BR4 ; lanes 8-12, A.
brasilense isolates ATCC 21145 (reference strain), AR8 , BR1 , CR3 , DR1 . Numbers indicate the size of some bands of the molecular mass
marker (100 bp ladder, Amersham biosciences).
fingerprints have been reported as a useful method to
identify Azospirillum strains of five different known
species (Grifoni et al., 1995; Han and New, 1998) and
also Bradyrhizobium strains (Vinuesa et al., 1998).
Obligate endophyte G. diazotrophicus was not iso-
lated from soil, rhizosphere and tissues samples col-
lected from four sugarcane fields in Granada area
(Spain) under conditions applied and the growth
medium used. Thus, isolates in semi-solid LGIP
medium (containing 0.5% of cane juice) showed an
slow growth after 5 days of incubation. In addi-
tion, colonies formed grew near to the surface of
the medium but did not form typical yellow pellicle
which is characteristic of G. diazotrophicus. Individ-
ual colonies pick up from LGIP agar plates and re-
streaked on semi-solid medium showed a poor growth
and did not express acetylene reduction activity after
successive transfers. These isolates were subcultured
and pure colonies were not identified as G. diazotroph-
icus based on physiological and biochemical tests
described in material and methods.
Related to this, our isolates were able to grow
in LGIP medium containing malate as carbon source
meanwhile the growth of the reference strain PAL-5
in this medium was unsuccessful. Furthermore, these
isolates did not grow on LGIP medium with sucrose
30% and not dark brown colonies were formed on
potato P medium. On the other hand, isolates and
extracts of sugarcane tissues did not produce a band
411-bp characteristic of G. diazotrophicus when sub-
jected to PCR reaction with primers 1440 and AD.
Detection of amplified PCR products was possible
from cell cultures of PAL-5 strain and sugarcane plants
229
inoculated with this strain cultured in growth cham-
ber (data not shown). Thus, our data suggest that G.
diazotrophicus populations in varieties of sugarcane
used in our study might be low in soil, rhizosphere
and sugarcane so they can not be isolated and detected
with the procedure used, and in this case their con-
tribution to nitrogen nutrition of this crop might be
poor.
High doses of N-fertilisers (400–500 kg N ha−1
−1
yr ) applied for the farmers on sugarcane crop could
be one of the reasons for the failure on the detection
and isolation of G. diazotrophicus from Spanish sugar-
cane varieties studied. As has been previously reported
by dos Reis Junior et al. (2000), Fuentes-Ramirez
et al. (1993) and Muthukumarasami et al. (2002), the
population of G. diazotrophicus was sensitive to the
application of N-fertilisers. We do not discard other
factors, such as varietal (Muthukumarasami et al.,
1999) and environmental effects (Bellone et al., 1996,
dos Reis Junior et al., 2000) that could be responsi-
ble for the results commented above. Samples from
sugarcane fields of our experiment were taken dur-
ing May and in these typical Mediterranean conditions
rainfall were low 42 l m−2 . On the other hand, Ortega
et al. (2001) has suggested that in the apical stem re-
gion (the same samples collected in our experiment)
the population of G. diazotrophicus should be higher
compared to basal and intermediate regions. In ad-
dition, the dynamic population of G. diazotrophicus
seem to be higher in first stages of the crop (3 month
old) and decrease drastically in relation to plant age
at least in Mexican sugarcane cultivars (Muñoz-Rojas
and Caballero-Mellado, 2003).
230
Figure 3. Agarose gel electrophoresis of amplified 16S rDNA of five A. chroococcum and four A. brasilense isolates digested with restriction
Nde II and Taq I. Lanes 1–7, A. chroococcum isolates CECT 4435 (reference strain), AS1 , AS2 , BS4 , AR2 , BR3 , BR4 ; lanes 8-12, A. brasilense
isolates ATCC 21145 (reference strain), AR8 , BR1 , CR3 , DR1 . Numbers indicate the size of some bands of the molecular mass marker (100
bp ladder, Amersham biosciences).
We have isolated from root surface of sugarcane
soil bacteria which rapidly grew on N-free media and
showed acetylene reducing activities similar to ref-
erence strains of Azotobacter spp. and Azospirillum
spp. growing in laboratory cultures. According to
biochemical, morphological and molecular tests, iso-
lates were identified as Azotobacter chroococcum and
Azospirillum brasilense. However, A. lipoferum and
obligate endophyte G. diazotrophicus were not found
in sugarcane associated microbial populations.
For the last two decades, different nitrogen fix-
ers bacteria have been considered the most promising
plant-growth promoting rhizobacteria. Azospirillum is
able to colonize in large numbers the roots of agro-
nomically important crop plants, grasses and forest
trees, from diverse regions of the world (Döbereiner
and Pedrosa, 1987). Thus, it could be suggested that
in natural environments, where sugarcane plants are
not under nitrogen stress, the production of stimu-
latory factors by PGPR like Azospirillum could be
considered beneficial for sugarcane plants.
Acknowledgements
We are grateful to Dr. José Herrera-Cervera for help-
ful comments. This work was supported by a grant
from Agencia Española de Cooperación Internacional
(AECI), Spain.
References
Baker F J 1967 Handbook of Bacteriological Technique. 2nd, ed.,
pp. 42. Butterworth, London.
Baldani J I, Caruso L, Baldani V L D, Goi S R and Döbereiner J
1997 Recent advances in BNF with non-legume plants. Soil Biol.
Biochem. 29, 911–922.
Becking J H 1992 The family Azobacteraceae. In The Prokaryotes,
A Handbook on the Biology of Bacteria: Ecophysiology, Isola-
tion, Identification, Applications. Eds. A Balows, G H Trüper, M
Dworkin, W Hander and K H Schleifer. Vol. 4, pp. 3144–3170.
Springer, Berlin Heidelberg, New York.
Bellone C H, Bellone S C and Pedrasa S C 1996 Hydraulic de-
ficiency and acethylene reduction in sugar cane roots. In 7th
Int. Symp. Nitrogen Fixation with Non-Legumes, 16 – 21 Oc-
tober. Eds. K A Malik, M S Mirza and A M Khalid. Faisalabad,
Pakistan, pp. 125–126.
Bergey’s Manual of Determinative Bacteriology 9th 1994 Eds. P H
A Sneath, J T Staley and S T Williams. Williams and Wilkins,
Baltimore, USA.
Bergey’s Manual of Systematic Bacteriology 1984 Eds. S T
Williams, M E Sharpe and J G Holt. Williams and Wilkins,
Baltimore, USA.
Boddey R M, de Oliveira O C, Urquiaga S, Reis V M, Olivares F
L, Baldani V L D and Döbereiner J 1995 Biological nitrogen
fixation associated with sugar cane and rice: Contributions and
prospects for improvement. Plant Soil 174, 195–209.
Boddey R M, Urquiaga S, Reis V and Döbereiner J 1991 Biological
nitrogen fixation associated with sugarcane. Plant Soil 137, 111–
117.
Caballero-Mellado J, Fuentes-Ramirez L E, Reis V M and Martínez
Romero E 1995 Genetic structure of Acetobacter diazotrophicus
populations and identification of a new genetically distant group.
Appl. Environ. Microbiol. 61, 3008–3013.
Cavalcante V A and Döbereiner J 1988 A new acid-tolerant bac-
terium associated with sugarcane. Plant Soil 108, 23–31.

de Bruijn F J 1992 Use of repetitive (repetitive extragenic palin-
dromic and enterobacterial repetitive intergeneric consensus)
sequences and the polymerase chain reaction to fingerprint the
genomes of Rhizobium meliloti isolates and other soil bacteria.
Appl. Environ. Microbiol. 58, 2180–2187.
Döbereiner J and Pedrosa F O 1987 Nitrogen-fixing Bacteria in Non
Leguminous Crop Plants. Science Tech. Publishers. Springer
Verlag Inc., Berlin, Germany.
Döbereiner J, Urquiaga S and Boddey R M 1995 Alternatives for
nitrogen nutrition of crops on tropical agriculture. Fertil. Res.
42, 339–346.
Dong Z, Canny M J, McCully M, Reboredo M R, Fernández C,
Ortega E and Rodés R 1994 A nitrogen-fixing endophyte of sug-
arcane stems: A new role for the apoplast. Plant Physiol. 105,
1139–1147.
dos Reis Junior F B, Reis V M, Urquiaga S and Döbereiner J
2000 Influence of nitrogen fertilisation on the population of
diazotrophic bacteria Herbaspirillum spp. and Acetobacter di-
azotrophicus in sugar cane (Saccharum spp.). Plant Soil 219,
153–159.
Fuentes-Ramírez L E, Caballero-Mellado J, Sepúlveda J and
Martínez-Romero E 1999 Colonization of sugarcane by Aceto-
bacter diazotrophicus is inhibited by high N-fertilization. FEMS
Microbiol. Ecol. 29, 117–128.
Fuentes-Ramírez L E, Jiménez-Salgado T, Abarca-Ocampo I R
and Caballero-Mellado J 1993 Acetobacter diazotrophicus, an
indoleacetic acid producing bacterium isolated from sugarcane
cultivars of Mexico. Plant Soil 154, 145–150.
Gómez F, Salmeron V, Rodelas B, Martínez-Toledo M V and
González-López J 1998 Response of Azospirillum brasilense to
the pesticides bromopropylate and methidation on chemically
defined media and dialysed-soil media. Ecotoxicology 7, 43–47.
González-López J 1992 Microorganismos diazotrofos asociados
a raíces de plantas no leguminosas. In Interacción Planta-
Microorganismo: Biología del Nitrógeno. Eds. J González-
López and C Lluch. pp. 71–96. Rueda, Madrid, Spain.
Grifoni A, Bazzicalupo M, Di Serio C, Farcelli S and Fani R 1995
Identification of Azospirillum strains by restriction fragment
length polymorphism of 16S rDNA and the histidine operon.
FEMS Microbiol. Lett. 127, 85–91.
Haahtela K, Wartiovaara T, Sundman V and Skujins J 1981 Root
associated N2 fixation (acetylene reduction) by Enterobacteri-
aceae and Azospirillum strains in cold-climate spodosoils. Appl.
Environ. Microbiol. 41, 203–206.
Han S O and New P B 1998 Variation in nitrogen fixing abil-
ity among natural isolates of Azospirillum. Microb. Ecol. 36,
193–201.
Hayward A 1964 Characteristics of Pseudomonas solanaceaerum.
J. Appl. Bacteriol. 27, 265–277.
Herrera-Cervera J, Caballero-Mellado J, Laguerre G, Tichy H,
Requena N, Amarger N, Martínez-Romero E, Olivares J and
Sanjuan J 1999 At least five rhizobial species nodulate Phaseolus
vulgaris in a Spanish soil. FEMS Microbiol. Ecol. 30, 87–97.
Kirchhof G, Baldani J I, Reis V M and Hartmann A 1998 Molec-
ular assay to identify Acetobacter diazotrophicus and detect its
occurrence in plant tissues. Can. J. Microbiol. 44, 12–19.
Kirchhof G, Reis V M, Baldani J I, Eckert B, Döbereiner J and
Hartmann A 1997 Occurrence, physiological and molecular
analysis of endophytic diazotrophic bacteria in graminaceous
energy plants. Plant Soil 194, 45–55.
Krieg N R and Holt J G 1984 Bergey’s Manual of Systematic Bac-
teriology Vol. 1, pp. 94. Williams & Wilkins, Baltimore-London.
Laguerre G, Allard M R, Revoy F and Amarguer N 1994 Rapid
identification of rhizobia by restriction fragment length poly-
231
morphism analysis of PCR-amplified 16S rRNA genes. Appl.
Environ. Microbiol. 60, 56–63.
Li R-P and McRae I C 1991 Specific association of dia-
zotrophic acetobacters with sugarcane. Soil Biol. Biochem. 23,
999–102.
Linblad P and Guerrero M G 1993 Nitrogen fixation and nitrate
reduction. In Photosynthesis and Production in a Changing En-
vironment. A Field and Laboratory Manual. Eds. D O Hall, J M
O Scurlock, H R Bolhar-Nordenkampf, R C Leegood and S P
Long. pp. 444–464. Chapman and Hall, London.
Martínez-Toledo M V, González-López J, de la Rubia T and Ramos-
Cormenzana A 1985 Isolation and characterization of Azotobac-
ter chroococcum from the roots of Zea mays. FEMS Microbiol.
Ecol. 31, 197–203.
Muñoz-Rojas J and Caballero-Mellado J 2003 Population dynamics
Gluconacetobacter diazotrophicus in sugarcane cultivars and its
effect on plant growth. Microbial Ecol. 46, 454–464.
Muthukumarasamy R, Revathi G and Lakshminarasimhan C 1999
Influence of N fertilisation on the isolation of Acetobacter di-
azotrophicus and Herbaspirillum spp. From Indian sugarcane
varieties. Biol. Fert. Soils 29, 157–164.
Muthukumarasamy R, Revathi G and Loganathan P 2002 Effect of
inorganic N on the population, in vitro colonization and mor-
phology of Acetobacter diazotrophicus (syn. Gluconacetobacter
diazotrophicus). Plant Soil 243, 91–102.
Oak A 1992 A re-evaluation of nitrogen assimilation in roots.
BioSci. 42, 103–111.
Olivares F L, Baldani V L D, Reis V M, Baldani J I and Döbereiner
J 1996 Occurrence of the endophytic diazotrophs Herbaspirillum
spp. in roots, steams and leaves predominantly of gramineae.
Biol. Fertil. Soils 2, 197–200.
Ortega E, Rodés R, de la Fuente E and Fernádez L 2001 Does the
routine treatment of sugarcane stem pieces for xylem pathogen
control affect the nitrogenase activity of an N2 -fixing endophyte
in the cane? Aust. J. Plant Physiol. 28, 907–912.
Reis V M, Olivares F L and Döbereiner J 1994 Improved methodol-
ogy for isolation of Acetobacter diazotrophicus and confirmation
of its habitat. World J. Microbiol. Biotechnol. 10, 101–104.
Rhodes M E 1958 The cytology of Pseudomonas spp. as revealed by
a silver-plating staining method. J. Gen. Microbiol. 18, 639–648.
Rodríguez-Cáceres E A 1982 Improved medium for isolation of
Azospirillum spp. Appl. Environ. Microbiol. 44, 990–991.
Socolofsky M D and Wyss O 1961 Cysts of Azotobacter. J. Bacte-
riol. 81, 946–954.
Soil Conservation Service 1975 Soil taxonomy. A basic system of
soil classification for making and interpreting soil surveys. US
Department of Agriculture, Washington DC, USA. 436 p.
Sprent J I and James E K 1995 N2 -fixation by endophytic bacte-
ria: questions of entry and operation. In Azospirillum VI and
Related Microorganisms. Eds. I Fendrik, M del Gallo, J Vander-
leyden and M de Zamaroczy. pp. 15–30. Springer-Verlag, Berlin,
Germany.
Stephan M P, Oliveira M, Texeira K R S, Martínez-Drets G and
Döbereiner J 1991 Physiology and dinitrogen fixation of Aceto-
bacter diazotrophicus. FEMS Microbiol. Lett. 77, 67–72.
Triplett E W 1996 Diazotrophic endophytes: Progress and prospect
for nitrogen fixation in monocots. Plant Soil 186, 29–38.
Ureta A, Alvarez B, Ramon A, Vera M A and Martinez-Drets G
1995 Identification of Acetobacter diazotrophicus, Herbaspir-
illum seropedicae and Herbaspirillum rubrisubalbicans using
biochemical and genetic criteria. Plant Soil 172, 271–277.
Urquiaga S, Cruz K H S and Boddey R M 1992 Contribution of
nitrogen fixation to sugarcane: 15 N balance estimates. Soil Sci.
Soc. Am. J. 56, 105–114.
232
Vela G R and Wyss O 1964 Improved stain for visualization of
Azotobacter encystment. J. Bacteriol. 67, 476–477.
Vinuesa P, Rademaker J L W, de Bruijn F J and Werner D 1998 Gen-
otypic characterization of Bradyrhizobium strains nodulating en-
demic woody legumes of the Canary islands by PCR-restriction
fragment length polymorphism analysis of genes encoding 16S
rDNA (16S rDNA) and 16S-23S rDNA intergenic spacers, re-
petitive extragenic palindromic PCR gemomic fingerprint, and
partial 16S rDNA sequencing. Appl. Environ. Microbiol. 64,
2096–2104.
Yoneyama T, Muraoka T, Kim T H, Decaney E V and Nakanishi Y
1997 The natural 15 N abundance of sugarcane and neighbouring
plants in Brazil, the Philippines and Miyako (Japan). Plant Soil
189, 239–244.
Section editor: J.K. Vessey