Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (20.34 MB, 413 trang )
4
CARLOS A . NEYRA AND J. DOBEREINER
(Beijerinck, 1925; Schroder, 1932; Krasil'nikov, 1968; Dobereiner, 1966; AbdEl-Malek, 1971; and many others). Various anaerobic, facultative, and aerobic
bacteria are capable of fixing nitrogen in soil, in the rhizosphere, and in roots.
Interest in aerobic organisms has been generally greater because aerobic metabolism is more efficient, and because agricultural soils, with some exceptions (e.g.,
paddy rice), are well aerated. In most systems the availabitity of energy and
carbon substrates represents the major limiting factors to biological N2 fixation.
Nitrogen fixation of importance in soils has only been demonstrated after the
addition of carbon substrates (Mishustin, 1970; Brouzes et al., 1971; Abd-ElMalek, 1971) or when growing plants release part of their photosynthates
(Dobereiner and Alvahydo, 1959; Dobereiner, 1961; Dobereiner et ul., 1972a;
Day et ul, 1975b). In addition, plant root exudates can play an important role
in the establishment and maintenance of the rhizosphere population (Rovira,
1965b).
A good example of solar energy utilization for N2 fixation is the legume
symbiosis, where the energy requirement for nitrogen fixation is equivalent to
the requirement for nitrate reduction (Minchin and Pate, 1973; Gibson, 1976).
However, photosynthate availability is still considered a major limiting factor for
N2 fixation in soybeans (Quebedeaux et ul., 1975).
On the other hand some tropical grasses can grow and produce constant yields
without addition of nitrogen fertilizer to the soils, and it was suspected for many
years that substantial Nz fixation occurred in these systems (Parker, 1957;
Moore, 1966; Dobereiner, 1966). Because of their photosynthetic characteristics
(see Section I), most of these plants are in a favorable position with regard to
photosynthate availability for growth and N2 fixation. In the last 5 years the
evidence for Nz fixation in grasses has accumulated rapidly. The results obtained
by several authors for field-grown tropical plants are summarized in Table 11.
Although the measurement of uptake of "N-enriched N z represents the most
satisfactory method for evaluation of N2 fixation, the introduction of the
acetylene reduction method has represented a major breakthrough in the evaluation of N, fixation both in the laboratory and under field conditions. The work
of Schollhorn and Burris (1967) and Dillworth (1966) suggested that the rate of
acetylene reduction may be used as an index of the rate of Nz fixation. The
reduction of acetylene to ethylene (C, H2-C2 H4) and the measurement of
ethylene by gas chromatography has been extensively used for the assessment of
N2 fixation in grass-bacteria associations. The reader is referred to the literature
for further details on the use of "N as a tracer and for the acetylene reduction
method (Burris and Wilson, 1957; Stewart et uL, 1967; Hardy et al., 1968, 1973;
Burris, 1972, 1974; Dart et ul., 1972).
Several procedures for the assessment of N2 fixation by acetylene reduction
have been adopted. A general procedure for assaying excised roots was described
NITROGEN FIXATION IN GRASSES
5
for Paspalurn notaturn (Dobereiner et al., 1972a) and Dig'taria decurnbens
(Abrantes et al., 1975), and the same procedure with slight modifications has
been used for several other forage grasses and grain crops (von Biilow and
Dobereiner, 1975; Day et al., 1975b; van Berkum and Neyra, 1976; Sloger and
Owens, 1976). Steel cylinders of small diameter are very useful for taking cores
of small grasses from the field (Day et al., 1975a; Abrantes et al., 1975). A
variety of devices have also been described for in situ measurements of N2
fixation under field conditions (Balandreau et al., 1974; Balandreau, 1975;
Watanabe and Kuk-Ki-Lee, 1975).
A. C-4 GRASS SYSTEMS
I . Paspalurn notatum
The first tropical C-4 grass-bacterial association to be studied in detail was
that of Paspalurn notaturn-Azotobacter paspali. Five ecotypes in this grass
(tetraploid types) show a very specific association with Azotobacter paspali
(Dobereiner, 1966; Dobereiner and Campelo, 1971). Of the 33 ecotypes or
cultivars studied, only five (tetraploid types) stimulated A. paspali growth in the
rhzosphere. Establishment of the bacteria on the roots takes several months and
inoculation does not accelerate this rhizosphere association (Dobereiner and
Campelo, 1971). Field plants, transplanted with adhering soil into vermiculite
and watered with nitrogen-free nutrient solution, fixed 80 mg N per pot in 2
months, the amount necessary for normal growth (Dobereiner and Day, 1975).
CzH2 reduction assays with intact soil plant cores correlated well with excised
roots extracted from the soil and assayed after overnight preincubation in low
p 0 2 (Dobereiner et al., 1972a). Paspalurn notatum grown in sand, from seeds,
did not show A. paspafi establishment, except when glucose was added (Kass et
al., 1971). It is possible that besides A. paspali other microorganisms (e.g.,
mycorrhizal fungi) may be involved in the establishment of the association.
Mosse (1972) observed very intensive mycorrhizal infection of this grass. Inoculation of irradiated Brazilian soils with Endogone spores resulted in large
increases in forage yield in Paspalurn notaturn (Mosse, 1972).
Localization of A. paspali has been suggested to be in the mucagel layer
outside the root (Dobereiner et al., 1972a). The correlation of root piece
nitrogenase activity and enrichment culture activity in A . paspali sucrose
medium was highly significant (r = .08l) (Dobereiner and Day, 1974) when the
same root pieces were used for inoculation of the enrichment medium. Estimates
of nitrogen fixation in intact soil plant cores (10 cm #) by the CzH2 reduction
method were calculated to be 340 g N/ha per day. "N2 assays in smaller vessels
extrapolated to 110 g N/ha per day (calculated from data by De-Polli, 1976).
6
CARLOS A. NEYRA AND J. DOBEREINER
2. Sugar Cane
In many parts of the world this crop has been grown in monoculture for more
than 100 years without addition of nitrogen fertilizer and a survey in SBo Paulo
(Brazil) revealed that only half of the fields with this crop responded to nitrogen
fertilizer even if PK was also supplied (Verdade, 1967). Selective stimulation of
the nitrogen-fixing Beijerinckia under sugar cane vegetation and positive rhizosphere effects have been shown (Dobereiner, 196 1). Assays with the acetylene
reduction method indicate that in this crop only a minor part of the N2 is fixed
in or on the roots and most of it in the rhizosphere or in the soil (Dobereiner et
al., 1972b; Ruschel, 1976). Rain water can carry leaf exudates into the soil
which enhance Beijerinckia growth (Dobereiner and Alvahydo, 1959). Maximal
soil nitrogenase activities were found in rhizosphere soil and between the rows,
where the canopy closes (Dobereiner et al., 1972b). Sugar cane seedlings exposed to 's N2 indicated fixation, incorporation, and translocation of nitrogen
to the leaves (Ruschel et al., 1976). Spirillum lipofem8mdoes not appear to be
stimulated in the sugar cane rhizosphere (Dobereiner, 1976a) and this supports
the prevalence of Beijerinckia spp. as the major N2 fixer in this plant.
3. Digitaria decumbens
This grass contains several cultivars of agronomic importance, e.g., Pangola,
Transvala, and Slenderstem. These three grasses were grown in our experimental
fields from November 1973 to May 1975 (two summers, one winter) and
showed mean nitrogen yields of 1S O , 1.48, and 1.40 kg/ha per day, respectively.
The nitrogen gain of the soil (0-20 cm depth) calculated from Kjeldahl analyses
before and after this period was 405, 216, and 468 g/ha per day, respectively
(Schank et al., 1975). Intact soil plant core assays in the summer 1975 showed
nitrogenase activities equivalent to 880, 480, and 970 g N/ha per day, respectively (Day and Dobereiner, unpublished data). Similar values (1460 ? 85 g
N/ha per day for Transvala and 1326 g N/ha per day for Slenderstem) have been
estimated from the data of De-Polli (1976). In lsN2 experiments significant
incorporation and translocation was shown in both species (Table I).
The N2 -fixing bacteria most commonly associated with Digitaria decumbens is
Spirrillum lipofemm. In several experiments, significant correlations of root
piece nitrogenase activity with S. lipofemm enrichment culture activity were
found, suggesting that S. lipofemm is the major organism responsible for
nitrogenase activity on roots (Dobereiner and Day, 1976). The most active root
pieces showed strongly reducing sites within the cortex, where cells packed with
tetrazolium-reducing bacteria were found. Inactive root pieces did not show such
sites (Dobereiner and Day, 1976).
TABLE I1
Potential of N, Fixation in Field-Grown Tropical Forage Grasses Associated with N, -Fixing Bacteria
N, -ase activity
C, H, /h/g
Plant species
Andropogen gayanus (C, )
Aizdropogen spp. (C,)
Brachiaria mutica (C,)
B. rugulosa (C,)
B. brachylopha (C,)
Bulbostylis aphylanthoides
Cynoden dactilon (C,)
Cynoden dactilon (C, )
Cyperus rotundus (C, )
Cypents sp. (?)
Cyperus obtusiflorus (?)
Digitaria decumbens (C,)
Hyparrhenia rufa (C,)
Hyparrhenia rufa (C,)
Hyparrhenia dissoluta (?)
Melinis minutiflora (C,)
Panicum maximum (C, j
Panicum maximum (C,)
Paspalum notatum (C,)
Paspalum comersenii (?)
Pennisentm purpureum (C, 1
Pennisetum purpureum ( C , )
Country
Roots
Soil
References
Nigeria
Ivory Coast
Brazil
Brazil
Ivory Coast
Ivory Coast
Brazil
Nigeria
Brazil
Nigeria
Ivory Coast
Brazil
Brazil
Nigeria
Ivory Coast
Brazil
Brazil
Nigeria
Brazil
Nigeria
Brazil
Nigeria
15-270
50-380
150-750
5-150
100-140
74
11-269
10- 50
-
Day and Dart (personal communication)
Balandreau e t al. (1973)
Dobereiner and Day (1975)
Dobereiner and Day (1975)
Balandreau et al. (1973)
Balandreau et al. (1973)
nobereiner and Day (1975)
Day and Dart (personal communication)
Dobereiner et al. (1975)
Day and Dart (personal communication)
Balandreauet al. (1973)
Dobereiner and Day (1975)
Dobereiner and Day (1975)
Day and Dart (personal communication)
Balandreau e t al. (1 973)
Dobereiner and Day (1975)
Dobereiner and Day (1975)
Day and Dark (personal communication)
Dobereiner and Day (1975)
Day and Dart (personal communication)
Dobereiner and Day (1975)
Day and Dart (personal communication)
-
0
-
0-0.07
-
10- 30
-
2
30-620
21-404
20- 30
30-140
10- 15
13- 41
20-299
75
2-283
25-30
5-954
60
-
nMinimal and maximal values obtained with excised preincubated roots.
-
0-0.35
0-0.15
-
04.19
04.15
-
04.33
-
04.09
-
8
CARLOS A. NEYRA AND J. DOBEREINER
4. Other Forage Grasses
Excised root assays have shown that several other tropical C-4 forage grasses
are able to fur N2 (Table 11). In a 3-year experiment in Nigeria, soil under fallow
of Panicum maximum contained 0.18% N (15 cm depth) while fallows under
legumes (Leucena glauca and Cajanus cajan) contained only 0.13%. This difference corresponds to 250 kg N/ha per year (Greenland, 1975). This illustrates
the tremendous importance of such a fallow crop for the nitrogen balance of
tropical soils even if only part of this amount was due to bic Jgical N2 fixation
and the remaining to prevention from leaching or denitrification. Only limited
results from core assays are available. Balandreau and Villemin (1973) estimated
N2 fnation (C2 H,) rates of 10-1 5 kg N/ha per year (in situ assays) in Ivory Coast
savannas where Panicum maximum and Andropogon sp. were predominant.
These authors found N,-fixing aerobes to be predominant in the rhizosphere but
did not identify them or relate them specifically to root nitrogenase activity. A
survey of S. lipoferum occurrence in various parts of Brazil revealed a high
incidence of this organism where Panicum maximum replaced virgin forest
(Diibereiner et al., 1976). In another experiment (six sites, 10 samples each)
there was a significant difference in S. lipoferum incidence between forage
grasses. Panicum maximum and Brachiaria mutica were the most favorable and
Hypawhenia rufa the least (Dobereiner, 1976b).
The mode of infection of Panicum maximum by Spirillum lipoferum has been
investigated by electron microscopy in axenic seedlings (Garcia et al., 1976). The
bacteria were observed on the root surface within 24 hours and in the middle
lamellae of the root cells within a week. No intracellular infection was observed
even after 1 month. These authors have suggested that S. lipoferum enters the
roots with the aid of pectolytic enzymes.
5. Grain Crops
Maize and sorghum represent two of the major grain crops in the world. High
nitrogenase activities (up to 9000 nmoles CzH4/g roots per hour) were found on
excised, preincubated maize and sorghum roots in a lowland soil in Rio de
Janeiro State (von Bulow and Dobereiner, 1975). Other estimates by this
method range between 100 and 2000 nmoles CzH4/g roots per hour (von Bulow
and Dobereiner, 1975; Abrantes et nl., 1976; Barber et al., 1976; Okon et al.,
1977a; Sloger and Owens, 1976). However, very low or no activities were
reported from soil plant core and in situ assays (Balandreau and Dommergues,
1973; Barber et a l , 1976; Burris, 1976; Tjepkema and van Berkum, personal
communication; for discussion on this discrepancy see Section V).
In Rio de Janeiro (von Bulow and Dobereiner, 1975), Brasilia and Londrina
(Peres, Nery, and Dobereiner, unpublished data) Spirillum lipoferum was found
NITROGEN FIXATION IN GRASSES
9
to be abundant in all Nz-fixing maize and sorghum roots examined. Sloger and
Owens (1976) also report isolation of this organism from maize roots grown in
Beltsville, Maryland, while it was not found in Wisconsin or Oregon (Burris et al.,
1976; Barber et al., 1976). Field-grown maize plants in Wisconsin inoculated
with strains of S. lipofemm isolated from Digitaria roots in Brazil, showed
establishment of the bacteria inside the roots (Burris, 1976; Dobereiner et al.,
1976). Inoculated plants showed higher nitrogenase activity than uninoculated
ones, while nitrogen-fertilized plants had no activity (Burris et al., 1976; Barber
et al., 1976). The total number of bacteria in surface-sterilized maize roots was
similar t o the number of S. lipofemm in the inoculated maize roots (Okon et al.,
1976a).
Significant correlations (p = 0.01) between maize root piece activities and
enrichment culture activities were only obtained when the roots were previously
surface-sterilized (von Bulow and Db'bereiner, 1975). Detailed studies on the
localization of Spirillum lipofemm in maize and sorghum roots are not yet
available. Maize plants grown in sterilized sand and soil collected in Wisconsin,
showed nitrogenase activities when inoculated with S. lipofemm. The organism
was reisolated from surface-sterilized roots (Burris et al., 1976). Effects on plant
growth and nitrogen incorporation however were not significant in these experiments. In Oregon attempts to isolate Nz-furing bacteria from maize plants
yielded Enterobacter cloacae (Raju et al., 1972).
B. C-3 GRASS SYSTEMS
1. Rice
There is little doubt as to the substantial contribution of biological Nz
fixation to the N economy of this most important grain crop. For instance, a
total of 23 rice crops, in an 11-year experiment at the International Rice
Research Institute in the Philippines, were obtained from a nonfertilized field
with no apparent decline in the nitrogen fertility of the soil. About 45 to 60 kg
N/ha per crop were removed through straw and grain (Watanabe and Kuk-KiLee, 1975). This represents a substantial amount of N which had to be replaced
in order to maintain the fertility level of the soil. Blue-green algae and photosynthetic bacteria account for a large part of the Nz fixation in paddy rice
(Watanabe and Kuk-Ki-Lee, 1975; Elnawamy, 1976). This subject has been
reviewed elsewhere (Stewart, 1976; Venkataraman, 1975).
Bacterial Nz f i a t i o n in intact rice cultures grown in test tubes has been shown
by Rinaudo et al. (1971) by Kjeldahl analyses and acetylene reduction. For the
latter method, plants were removed and assayed after 24 hours of preincubation
under anaerobic conditions. The results obtained by the two methods were in
10
CARLOS A. NEYRA AND J . DOBEREINER
good agreement. Excised root assays of field-grown rice roots (Yoshida, 1971a;
Yoshida and Ancajas, 1973) confirmed “rhizosphere N2 fixation.” Results from
intact soil plant systems in the field gave about 50 to 200 g N/ha per day at the
flowering stage, by the nonalgal component. The algae were separated by
removing the flooding water and assayed separately for N2 fixation (Watanabe,
1976). Balandreau (1975) reported that 25 to 30 kg N/ha can be fixed for the
growing season by the nonalgal component.
Bacterial counts indicate that Beijerinckia sp. and Enterobacter cloacae are the
most common N2-furing bacteria in the rhizosphere of rice (Yoshida, 1971b;
Balandreau, 1975). However, the methods used by these authors would not
reveal Spirillum lipoferum. When various types of roots were compared, mature
roots with many laterals were the most active ones (Hamad-Fares et aZ., 1976).
Such roots, surface-sterilized for 1 hour with 1% chloramin T yielded almost
pure cultures of an organism with properties resembling S. lipoferum (Diem et
al., 1976). However, most of the nitrogen fixation in the rice system has been
attributed to rhizosphere soil rather than roots themselves (Yoshida and Ancajas,
1973). Higher numbers of aerobic than of anaerobic N2-fixing bacteria in the
rhizosphere of rice were also found by Dommergues et al. (1973) and Watanabe
and Kuk-Ki-Lee (1975). Aerobic or microaerophilic N2-fixing bacteria were also
found to be prevalent in roots of a salt marsh grass (Patriquin, 1976). Methaneoxidizing bacteria which are able to fix Nz were also found in rice paddies. The
large amount of CH4 which can accumulate in these soils, should not be
overlooked as a potential carbon source for Nz fixation (De Bont et aZ., 1976).
However, O2 diffusion seems a limiting factor for this system. Inhibition of CH4
oxidation by C2H2 and consequent interference in C2 H2 reduction complicate
estimates of Nz fixation where these organisms are present (De Bont and
Mulder, 1976). Very high numbers (up to 3.6 X lo7) of N2-fixingCH4-oxidizing
organisms were found in the rice rhizosphere (De Bont el al., 1976).
2. Wheal
A nitrogen balance study in the famous Broadbalk continuous wheat experiment carried out from 1843 to 1967 in England, showed an average annual gain
of 34 kg N/ha, of which 24 kg N/ha were removed with straw and grain (Jenkinson, 1973). However, values extrapolated from C2H2 reduction assays on cores
were much lower (2 to 3 kg N/ha per year) (Day et al., 1975a). It was also
shown that nitrogenase activity of soil cores containing wheat was significantly
higher than in bare soil (Day et al., 1975a).
Wheat cores assayed in Oregon have been calculated to fm 2 g N/ha per day
(Barger et al., 1976). Much higher nitrogenase activities have been observed in
wheat cores assayed in Rio de Janeiro (Table 111). Similar results were obtained
with cores from several wheat cultivars grown in pots in Parana (Brazil) (New
and Abrantes, personal communication). Excised root assays underestimated, by
11
NITROGEN FIXATION IN GRASSES
TABLE 111
Nitrogenase Activity in 10 Intact Wheat Cores (cv. Sonora) Collected at Random in the
Field, at Flowering Stage
Mean 2 most active cores
Mean 5 intermediate cores
Mean 3 least active cores
Mean all cores
nmol
C, H, /hour/core
g N, /day
10 cm@core
2641
1137
180
597 x
276 X
44 x 10-6
-
-
g N, /daylhaa
506
238
38
229
aEstimate by the theoretical C, H, :N, 3: 1 ratio from 24-hour rates based on the @ of
10-cm area of the cores corrected for 15-cm distance between rows.
about one-half, the core activities but showed significant correlations with the
core assays (r = 0.86 in Rio de Janeiro and r = 0.87 in Parana).
In the Broadbalk experiment a large part of Nz fixation was attributed to
blue-green algae but root nitrogenase activity was attributed t o anaerobic or
facultative bacteria (Day et al., 1975a). Barber el al. (1976) isolated N,,-fixing
strains of Enterobacter cloacae, Bacillus macerans, and B. polymyxa from wheat
roots in Oregon. On the other hand, enrichment cultures in semisolid N-free
malate medium inoculated with surface-sterilized wheat roots obtained from
different locations in Brazil (Rio de Janeiro, Parana, and Brasilia) yielded almost
100% positive samples of SpiriZlum lipofemm. Samples from R o Grande d o Sul
(extreme south of Brazil) showed that only 20% of the root samples were
positive for this organism. Attempts to correlate root piece nitrogenase activity
with enrichment culture activity in wheat have been unsuccessful.
Larson and Neal (1976) described a highly specific association of a facultative
Bacillus sp. with a disomic chromosome substitution line of wheat. The Bacillus
was isolated from a soil where wheat had been growing for 30 years without
nitrogen fertilizer. The rhlzosphere of this wheat line contained also more
nitrate-reducing bacteria and a lower total number of microorganisms. In
monoxenic culture, the bacterium closely associated itself with the root surface.
Abundant numbers of bacterial cells were found on the root surface as well as in
the intercellular spaces between the cortical root cells. Rovira (1965a) reported
establishment of an N,-fixing Bacillus sp. in wheat. Recent fine structure studies
by Foster and Rovira (1976) showed active penetration of wheat cortex cell
walls by bacteria, including Bacillus sp., at the flowering stage.
C. MISCELLANEOUS SYSTEMS
In addition to the N, -Axing systems previously described, which all bear some
relation to agricultural crops, a number of water plants and weeds have been
12
CARLOS A. NEYRA AND J. DOBEREINER
shown to exhibit substantial nitrogenase activity. An understanding of these
systems may help to clarify others of more immediate agricultural importance.
The tropical marine angiosperms K5alassia testudinum, Syringodium Jiliforme,
and Diplanthera wrightii and the temperate Zostera manna fixed an amount of
nitrogen sufficient for growth (Patriquin and Knowles, 1972). Thalassia testudinum Nz furation reached 100 to 500 kg N/ha per year. Conversion factors of
CzH2 reduction estimates as compared with estimates by lsNz incorporation
were close to the theoretical value 3 (2.6 to 4.6). A good correlation of numbers
of anaerobic Nz -furing bacteria and nitrogenase activity in glucose-amended
sediments was obtained, but aerobic N2 fixers were 50 to 300 times more
abundant in the rhizosphere than in the sediment and the authors concluded
that organisms other than Azotobacter and Clostridium are the predominant
nitrogen furers in these systems. Spartina altemiflora a C-4 grass from Canadian
salt marshes (Patriquin, 1976) was shown to have an association similar to that
described for Digifaria (Dobereiner and Day, 1976). Tetrazolium-reducing bacteria, similar to S. lipofemm, were found to be concentrated in the outer and
inner cortex layer of the roots. Nz-fixing aerobic bacteria resembling S. lipoferum were also isolated from Potamogeton Jilifonnis roots grown in Scottish
lakes (Silvester-Bradley, personal communication). Kgh nitrogenase activity has
alwo been observed in excised roots of mangroves (Rhizophora mangle and two
other species; Silver et al., 1976) and in intact soil plant cores of Juncus balticus
(Barber et al., 1976) and several inulin-containing plants (Dahlia pinnata and
others; Jain and Vlassak, 1975; Vlassak and Jain, 1976).
I I I . Bacteriology
Nitrogen-fixing bacteria which have been found in association with grasses are
all capable of fixing nitrogen in soil or culture medium without the plant.
Therefore, they are generally included in the group of “free-living Nz-fixing
bacteria” (Mulder and Brotonegoro, 1974). An excellent up-to-date review on
the entire group has been given by these authors and we will therefore restrict
this chapter to bacteria for which specific associations with tropical grasses have
been shown. Several species of Nz-fixing bacteria have been isolated from the
rhizosphere of temperate plants, e.g., the facultative Enterobacter cloacae, other
Enterobacter spp., and members of the Klebsiella aerobacter group (Raju et al.,
1972; Evans et al., 1972; Barber et al., 1976; Balandreau, 1975) but have not
been shown to be involved in nitrogenase activity on roots or in the rhizosphere.
A microaerophilic N2-fixing bacterium has been isolated from Digitmia sanguinalis with characteristics very similar to s. lipoferum (Barber and Evans,
1976).
The Azotobacter spp. (except A. paspali) are found mainly in the outer
rhizosphere of plants and can be very abundant under warm arid conditions
NITROGEN FIXATION IN GRASSES
13
(Vancura et al., 1965; Abd-El-Malek, 1971). Rice in Japan and India (Ishizawa
and Toyoda, 1964; Gopalakrishnamurthy et al., 1967) and sorghum in India
(Shantaram and Rangaswamy, 1967) have been shown to stimulate Azotobacter
growth. Books have been written on the presence of Azotobacter in the rhizosphere of plants (Rubenchik, 1963; Krasil’nikov, 1958) but, in more recent
reviews, Macura (1966), Mishustin (1970), and Rovira (1963) have come to the
overall conclusion that the growth of Azotobacter chroococcum and A. vinelandii is not influenced by plant roots. Even inoculation does not normally
establish these bacteria in the rhizosphere (Dobereiner, 1974). In some instances
establishment of Azotobacter on the roots has been observed, particularly when
the competitive rhizosphere microflora was eliminated (Riviere, 1959; Jackson
and Brown, 1966; for a more detailed discussion on this subject refer to
Diibereiner, 1974). Inoculation of Ammophyla arenaria (a sand grass) with
Azotobacter chroococcum had beneficial effects on plant growth and nitrogen
content (Abdel Wahab and Wareing, 1976). When rice seedlings in axenic
cultures were inoculated with Azotobacter chroococcum and two other bacterial
isolates from rice, acetylene reduction was observed. However, there was no
effect on plant growth. Even the highest activity observed (0.2 pg N per plant
per day) was too low t o meet the need for nitrogen (400 pg N in 3 weeks)
(Watanabe, 1975).
Inoculation of rice seedlings (grown in nonsterile soil) with Azotobacter or
Beijerinckia in France gave no increase in nitrogenase activity (MourarrCt et al.,
1975). However, inoculation with Spirillum lipoferum resulted in a threefold
increase in nitrogenase activity (Mourarrbt et al., 1975).
Anaerobic N2 -fixing bacteria are seldom stimulated in the rhizosphere. Still,
Katznelson ( 1 965) found that Clostridium numbers were increased in the rhizosphere of some plants and Rovira (1963) found that this organism could be
established in the root zone of wheat, maize, tomato, and lucerne.
The overall effect of a particular plant cover on the occurrence of certain bacteria is good evidence for meaningful rhizosphere or root associations. Currently,
information on such effects is available for only three tropical N2-fixing bacteria. These will be discussed in more detail.
A. Beijerinckia
Beijerinckia species are restricted to tropical and subtropical regions (Derx,
1953; Becking, 1961; Dobereiner, 1968). The acid tolerance of these organisms
enables them to compete successfully with most other soil microorganisms in
acid soils but they are unable to compete in neutral soils (Strijdom, 1967). In
sterile soils changes of pH had no effect.
A survey of Beijerinckia distribution in several Brazilian states demonstrated
the presence of this organism in 97% of the soil samples under sugar cane and
14
CARLOS A. NEY RA AND J. DOBEREINER
only 60% of the soil samples from other vegetations (Dobereiner, 1961). After
transplanting sugar cane to a new field, Beijerinckia numbers in the rhizosphere
and root surface soil increased steadily, but numbers of total bacteria, actinomyces, and fungi decreased. This suggests that changes in the microbial equilibrium of the sugar cane rhizosphere favor N2 -fixing Beijeiinckia.
Morphological and taxonomic characteristics of Beijerinckia species together
with methods for isolation and culturing have been described by several authors
(Becking, 1961; Dobereiner and Ruschel, 1958, 1964; Bergey, 1975). However,
very little is known about the physiology of this organism. There is no report on
the in vitro preparation of nitrogenase. Nitrogenase activity assayed in vivo
requires high levels of CzHz (40-80%) for saturation of the enzyme (Spiff and
Odu, 1973; Dobereiner, 1973). This organism is surrounded by a tough gum
which may act as a physical barrier to oxygen (or CzHz) and therefore function
as an oxygen protection mechanism for the nitrogenase, as suggested by Hill
(1971) for Derxiagummosa.
Mutants of Beijerinckia indica with less gum seem to be more sensitive t o
oxygen (Dobereiner, 1973). Very fast growth of Beijeiinckia can be obtained in
liquid media aerated with an N2 :Oz :COz (95:4.5:0.5) gas mixture but little
growth occurs without COz . Beijerinckia fluminensis does not form a gum but
shows characteristic zooglea-like clusters of four to eight organisms, surrounded
by a distinctive membrane. The function of this membrane is not known but it
has been suggested that it may retard oxygen diffusion and thereby protect the
nitrogenase (Dobereiner, 1974). The latter was not confirmed in chemostat
cultures with known p 0 2 .
Peiia and Dobereiner (1974) studied the effects of NO3- and NH4' on growth
and nitrogenase activity of Beijerinckia spp. Beijerinckia indica maintained about
30% of the original nitrogenase activity for 60 hours after the addition of 10 mM
N03- and 10-20% of the activity after the addition of 10 m M NH4'. In contrast,
Azotobacter vineIandii lost all nitrogenase activity in 3 to 3 hours after similar
additions of NO3- and NH4' (Peiia and Dobereiner, 1974). Beijerinckia fluminensis appeared to be unaffected by NO3- during a 6-hour incubation period.
Even after 3 days 50% of the nitrogenase activity was still present. The capacity
to fix NZ even in the presence of hgh concentrations of mineral nitrogen may be
of considerable ecological importance. These bacteria have not been examined
for a dissimilatory nitrate reductase.
B. A z o t o b a c t e r paspali
Azotobacter paspali, although capable of fxing N2 independently of the plant,
is ecologically restricted to the rhizosphere of a few ecotypes of Pasparum
notatum. This aerobic bacterium was found t o occur in 98% of the 252 root