Tải bản đầy đủ - 0 (trang)


Tải bản đầy đủ - 0trang



Inoculation with nitrogen-fixing microorganisms is known to cause nitrogen gains associated with the cultivation of legumes and sometimes cereals.

However, such inoculation often has only a small effect, at least compared

with the potential nitrogen gains that would occur if the microorganisms

were added to plants grown in sterile soil or to unplanted sterile flooded or

nonflooded soil. Often, the inoculum has no effect at all. In the case of

legumes, for example, inoculation with Rhizobium sometimes does not lead to

the appearance of nodules formed by the inoculum (Obaton, 1977), nodules

being required for nitrogen fixation by this symbiosis. In some cases, only

about 20% (Kuykendall and Weber, 1978) or as few as 5 % (Dobereiner,

1978) of the nodules are derived from the inoculum strain of Rhizobium, the

use of which is recommended because of the low nitrogen-fixation effectiveness of the indigenous rhizobia in the soil. Similarly, the poor ability of

Rhizobium to colonize the roots even of its host legumes is evident in the

benefit arising from large as compared to small inocula applied to clover

(Holland, 1970) and soybeans (Kapusta and Rouwenhorst, 1973).

Extensive colonization of the rhizosphere of legumes by effective strains of

Rhizobium, of the root zone of nonlegumes by free-living nitrogen-fixing

bacteria, and of flooded soils and the overlying water is necessary for nitrogen

gains to be extensive. This is true for both indigenous nitrogen fixers and

organisms used as inocula. In the case of Rhizobium, the degree of colonization when roots are susceptible to infection is reflected in the extent of

nodulation. Thus, a threshold number of the root-nodule bacteria appears to

be necessary for nodulation to occur (Bohlool and Schmidt, 1973; Purchase

and Nutman, 1957), and the abundance of nodules is directly related to the

number of rhizobia in the root zone (Lim, 1963). For the free-living bacteria

and blue-green algae, the extent of nitrogen accretion is likely a direct

function of the extent of growth because a small biomass probably brings

about little nitrogen input and growth is required for the initially small

population in the soil or the inoculum to give the large cell mass necessary for

significant activity. This colonization requires that the active species attain or

maintain large populations in environments with numerous competitors,

many of which grow faster, as well as protozoa that prey on them and

invertebrates that graze on them.

Despite the abundance and activity of competitors, predators, and other

organisms that may be inimical to the nitrogen fixers, few attempts have been

made to favor the inoculum organism in some selective way. The inoculum

strain is added to the seed, the floodwater, or sometimes the soil, and in the

absence of a means of selective stimulation, large inocula are used. Even

many of these large inocula are of little value. In contrast, the approach

suggested here is designed to overcome the natural biotic stresses or

biological controls that hold the nitrogen fixers in check. The approach

involves (1) first establishing the importance of predation, competition, or



other biotic stresses that suppress the desired organism and (2) then seeking

chemicals that inhibit the predators, competitors, or other antagonists.

Should the nitrogen-fixing organism whose establishment is desired be

naturally susceptible to the compounds, mutants or other cultures are

obtained that are resistant to the inhibitors. In this way the biotic stress is

reduced or overcome, and the introduced nitrogen fixer has a selective

advantage in the heterogeneous community in which its establishment is



Considerable attention has been given to explaining why populations of

Rhizobium are rarely large in soil, why they do not reach as large numbers in

the rhizosphere as species of other gram-negative bacteria, and why large

numbers experimentally added to soil or the high cell densities that briefly

appear in the rhizosphere decline markedly. The likely biological explanations include the deleterious actions of bacteriophages, lytic organisms,

antibiotic producers, competitors, Bdellovibrio, myxobacteria, or protozoa.

However, the density in soil of Bdellovibrio and bacteriophages able to

destroy Rhizobium appears to be too small to have a significant effect on the

population of root-nodule bacteria (Keya and Alexander, 1975; Ramirez and

Alexander, 1980). The numbers of myxobacteria and of Bdellovibrio, bacteriophages, and lytic microorganisms able to destroy Rhizobium juponicum

have been found to be fewer than 300 per gram of rhizosphere soil (Hossain

and Alexander, unpublished data), numbers that are probably too small to

significantly affect the R. juponicum population. Similarly, small numbers of

lytic microorganisms, Bdellovibrio, and bacteriophages were found in the

rhizosphere of Phaseolus vulgaris at the time that the population size of

Rhizobium phaseoli was declining by about three orders of magnitude

(Lennox and Alexander, 1981). Hence, the cause of the biologically induced

decline or biological control of the rhizobia must be sought elsewhere.

It has long been known that protozoa are more abundant in the root zone

than in adjacent nonrhizosphere soil (Katznelson, 1946). Bacteria flourish at

the expense of the array of readily metabolizable organic compounds

excreted by roots, and the resulting large cell densities are highly suitable as

prey for the protozoa that presumably are active when they have many cells

on which to feed. Direct evidence for the role of protozoa in reducing high cell

densities of Rhizobium was obtained in studies of nonrhizosphere soil to

which large rhizobial populations were deliberately added (Danso and

Alexander, 1975; Habte and Alexander, 1977). Subsequent research suggested that the same relationship held in the rhizosphere of P . vulgaris. Thus,

R. phuseoli inoculated onto bean seeds proliferated rapidly shortly after



planting the seeds, and the rhizobium soon became quite abundant. As the

rhizobia as well as other bacteria replicated, their high densities led to a

marked rise in the abundance of protozoa. The protozoa, in turn, brought

about a drastic reduction in the abundance of all bacteria, including R.

phaseoli (Lennox and Alexander, 1981).

Protozoa not only may markedly reduce the size of large populations but

they also may prevent the development of such large populations when the

cell density initially is low. This is suggested by observations that small

numbers of R. phaseoli, Rhizobium meliloti, R. japonicum, and a strain

nodulating cowpeas failed to increase in abundance in natural soil even when

a carbohydrate they can use was added, but they increased from 6- to 170fold in 4 days when inhibitors of eukaryotic microorganisms were added

(Pena-Cabriales and Alexander, 1983).The suppression by protozoa of small

numbers of rhizobia and the reduction in abundance of rhizobia to values

below about lo6 per gram of soil, which were sometimes observed (Chao and

Alexander, 198 l), were unexpected because it was generally believed that

protozoa could not actively graze on small populations of bacteria, that is,

populations below about lo6 per gram (Alexander, 198 1). However, although

the need for a threshold density of prey does indeed appear to be necessary

for such predation, not all prey cells need to be of the same species, and the

protozoa are able to feed on alternative species in addition to the bacteria of

interest. In this way, the protozoa graze actively and replicate at the expense

of the alternative prey, the density of which is above the threshold, but in their

relatively indiscriminate grazing, the protozoa also consume and thereby

affect a particular species existing at a value below the threshold (Mallory et

al., 1983). Such an effect of alternative prey is particularly likely in the

rhizosphere, in which enormous numbers of gram-negative bacteria are

present and are growing by using the supply of organic nutrients released by

the roots. Hence, a suppression by protozoa of the slow-growing rhizobia,

few of which can reach the population sizes of the many fast-growing

bacterial species in the root zone, is probably widespread.

The results of a test of the possible significance of alternative prey to the

suppression of Rhizobium provide further evidence for the ecological role of

protozoa. For example, when simple organic nutrients are added to soil to

simulate the products released by plant roots, bacteria able to use these

nutrients multiply. This leads to a dramatic rise in the population of

protozoa. At the same time, R. phaseoii declines to values at day 7 of less than

0.1% of the initial population size. Similarly, the deliberate addition to soil of

cells of an alternative prey for protozoa results in a suppression of R. phaseoli

(Ramirez and Alexander, 1980).

Granting that high rates of nitrogen fixation require the presence of large

numbers of an effective strain of Rhizobium when legume roots are susceptible to infection, and granting further that protozoa are important in



suppressing the rhizobia or holding them in check, then a chemical that is

toxic to protozoa but has no direct effect on rhizobia or their leguminous

host should enhance colonization by the root-nodule bacteria, promote

nodulation, and hopefully bring about greater nitrogen fixation. Initial

investigations to test these possibilities involved the use of mannitol-amended

soil to simulate root excretions and additions of cycloheximideand Triton X100, a mixture that is highly detrimental to protozoa. Under these conditions,

R. phaseoli proliferated extensively, rather than becoming more sparse

(Ramirez and Alexander, 1980). In subsequent studies, thiram [bis(dimethylthiocarbamoyl)disulfide], which is highly toxic to protozoa, was coated on

the seeds. The host plant was P. oulgaris, and a thiram-resistant strain of R.

phaseoli was inoculated directly onto seeds, as is the common field practice.

The rise in abundance of R. phaseoli was identical in the presence or absence

of the pesticide, but the subsequent fall in their abundance was appreciably

slowed by the chemical. Of special importance, the population of R. phaseoli

was higher on roots derived from thiram-treated seeds at the time that

rhizobial infections were being initiated on young roots and nodules were

being formed on the bean seedlings. Under the test conditions, thiram killed

many of the protozoa, and for several days their numbers remained smaller

than in the rhizosphere of bean seedlings derived from unamended seeds.

With time, however, the protozoan population increased. It was not clear

whether the return of the protozoa resulted from the breakdown of the

pesticide, the appearance of resistant protozoa, or protozoan colonization of

roots at a site distant from the point of placement of the treated seeds. That

the beneficial effect of thiram resulted from its inhibition of protozoa was

confirmed by a study in which it was found that stimulation of R. phaseoli

took place in sterile soil inoculated with a protozoa-containing mixture of

soil microorganisms, but not if the mixture was protozoa free (Lennox and

Alexander, 1981).

Similar effects have been noted in the rhizosphere of soybeans derived from

seeds coated with benomyl [methyl 1-(butylcarbamoyl)-2-benzimidazolecarbamate] and inoculated with a benomyl-resistant strain of R. japonicum. The

fungicide reduced the density of protozoa in the rhizosphere and slowed the

decline of R. japonicum after its initial flush of growth. The suppression of

protozoa by benomyl also disappeared, as occurred with thiram applied to

beans, as the soybeans developed further (Hossain and Alexander, unpublished data). By maintaining larger populations of rhizobia, the pesticides

presumably increased the opportunities for the bacteria to infect root hairs as

the root system developed, but the effect was limited to young plants.

From a practical viewpoint, the critical issue is not the inhibition of the

potential predator or the enhanced colonization of the rhizobia, but rather

the frequency of nodulation, the relative abundance of effective nodules, and

particularly the response of the above-ground portions of the plant. This



issue was first addressed in investigations of P . vulgaris. Application of thiram

and a thiram-resistant R. phuseoli strain to seeds increased the number of

nodules on 21-day-old plants as well as the percentage of nodules formed by

the resistant strain in a soil containing other R. phuseoli strains. The

differences in nodule numbers on beans derived from treated and untreated

seeds disappeared as the plants matured. Nevertheless, adding thiram to the

inoculated seeds increased the nitrogen content and yield of the tops of the

beans and also increased the number and weight of the pods as compared to

plants derived from inoculated seeds not treated with the fungicide. Furthermore, when the resistant rhizobia were present in the soil, as would occur if

these bacteria were introduced as a consequence of inoculation during a

preceding cropping season, pesticide treatment of the seed still increased the

yield and nitrogen content of the plants. The suppression of protozoa and the

consequent improved colonization by the effective, pesticide-resistant rhizobia appear to explain the beneficial responses of the plant (Lennox and

Alexander, 1981).

In trials with soybean, the seeds received benomyl or oxamyl (N',N'dimethyl-l\r-[(methylcarbamoyl)oxy]-1-thiooxamimidic acid methyl ester)

and R. juponicum strains resistant to one or the other of these pesticides. The

combination of benomyl and the bacteria increased the relative frequency of

nodules formed by the inoculated rhizobia and also increased the percentage

nitrogen, the total nitrogen content, the yield at one of the planting densities

tested, and the pod weight in one of the soils examined in the greenhouse. The

combination of resistant R. japonicum and oxamyl-which was applied to the

seeds, foliage, or both-increased the relative abundance of nodules produced by the inoculum organism and also increased the yield, percentage

nitrogen, total nitrogen content, and weight of pods and grain. The beneficial

effects on soybeans grown in the greenhouse could have resulted from a

favoring of the resistant strain of R.japonicum, a suppression of pathogens, or

both (Hossain and Alexander, unpublished data).

Predation appears to be the major, and possibly the sole, reason for the

decline of the large rhizobium populations that appear as seeds germinate

and seedlings develop. However, predation is not the sole mechanism

involved in preventing the increase in rhizobial numbers. The most numerous

microorganisms in the rhizosphere are bacteria, many of which undoubtedly

compete with rhizobia for the supply of organic compounds needed for

growth. Moreover, because many of these bacteria grow rapidly and some

require few or no growth factors, they probably are better competitors than

the slower growing and often more fastidious rhizobia. Such organisms

probably use up nearly all of the organic carbon that is excreted from the

roots, leaving little for the root-nodule bacteria. Although competition by

fast-growing species is a likely mechanism to prevent or reduce the extent of



growth of slow-growing organisms, it is not easy to demonstrate unequivocally its significance in natural ecosystems. Possibly the best way to establish

its importance is to eliminate all other likely mechanisms for the biological

control of a test species and then to show that suppression of the likely

competitors relieves the check maintained on growth of the population of

interest. This is the approach that has been used in an investigation of R.

phaseoli, R. rneliloti, R.japonicurn, and an isolate that nodulates cowpeas. The

soil contained too few bacteriophages or cells of Bdellovibrio or myxobacteria

to have an appreciable impact. Other studies failed to reveal significant

numbers of cells capable of producing lytic enzymes or antibiotics active

against rhizobia. In these circumstances, the addition to soil of streptomycin

and erythromycin, antibiotics that have little or no effect on protozoa but

that are toxic to bacteria (which are the most likely competitors with

rhizobia), resulted in an enormous stimulation of the four rhizobia. For the

purposes of the study, the rhizobia were made resistant to the two antibiotics.

The density of the rhizobia increased by 400- to 3700-fold, a stimulation far

greater than that observed when protozoa and other eukaryotes were

suppressed. In line with the suggestion that competition limits the growth of

small populations of Rhizobium in soil was the finding that their numbers

were not detectably increased by additions of 0.01 or 0.10% mannitol, but

that the counts rose upon the addition of 0.50% mannitol; that is, the

competitors used up the lower concentrations before sufficient time had

elapsed for the rhizobia to replicate to a significant degree, but enough of the

carbon source was available at the higher concentration for even the slowgrowing species (Pena-Cabriales and Alexander, 1983). Earlier studies also

showed that Rhizobium trifolii and Rhizobiurn lupini only increased in

number in soil if large amounts of carbohydrates were added (Chowdhury,


If the explanation of the effects of the antibiotics is correct, then the use of

antibiotics, synthetic chemicals, or other inhibitors that are detrimental to the

competitors should result in markedly enhanced growth of mutants of

Rhizobium that are tolerant of the inhibitors. If the chemicals have a broad

spectrum of action and are injurious to both competitors and predators, the

should be still more marked.

stimulation of Rh~zob~um


The approach of using inhibitors to control predators and competitors

together with the inhibitor-tolerant inocula has not been pursued with the

nitrogen-fixing bacteria that may enhance the growth of cereals or other



nonlegumes. Such bacteria sometimes bring about small nitrogen gains under

natural conditions (Balandreau and Villemin, 1973; Steyn and Delwiche,

1970; Vlassak et al., 1973), but their abundance is rarely great enough to

exploit fully the organic excretions of the root. To help overcome the scarcity

of nitrogen fixers, the practical use of inoculation has been investigated, e.g.,

Azospirillum sp. on millet (Barber et al., 1979) and Azotobacter chroococcum

(Thompson, 1974a) and Beijerinckia derxii (Thompson, 1974b) on wheat.

Although the amount of carbon excreted by the root may rarely be great

enough to support the fixation brought about within the nodule by Rhizobium, allowing the free-living nitrogen fixer to use more of the available

carbon than it otherwise would, but for the competition with other organisms

or the impact of predation, should result in greater nitrogen gains.

That the approach is feasible is evident from studies designed to promote

colonization of a strain of Pseudomonas in the corn rhizosphere (MendezCastro and Alexander, 1983). The fungicide used was mancozeb [Mn and Zn

salt of ethylenebis(dithiocarbamate)], and the isolate was a strain resistant to

this pesticide. Following inoculation of the pseudomonad onto corn seeds,

the bacterium became established on the roots, but its abundance was never

great. In contrast, when the seeds were coated with mancozeb at the time they

were inoculated, the roots that appeared had populations of the inoculum

strain that were 100-fold or more larger. The types of organisms suppressed

by mancozeb were not established.


In fields of rice planted in flooded soil, blue-green algae (cyanobacteria)

may add appreciable nitrogen by fixation (App et al., 1980; Roger and

Kulasooriya, 1980), and much or nearly all of the nitrogen removed from the

soil with the harvested grain may be returned by their activity. The quantity

of nitrogen fixed undoubtedly varies appreciably with the type of algae that

dominates in the flooded field and with the prevailing environmental

circumstances, but a figure of 30 kg/ha/yr seems reasonable under natural

conditions (Watanabe et a]., 1977). Because the indigenous algae may not be

active in fixation, inocula are sometimes added to the paddy water to increase

the microbial fixation of nitrogen, and these inoculations may sometimes

increase the yield of rice (Roger and Kulasooriya, 1980).

Several factors affect the rate and extent of development of these algae and

the amount of nitrogen they fix. Among the more significant of these factors

are grazing and competition by other organisms that proliferate in the



floodwater. As the algae multiply, the large biomass they create serves as a

food source for invertebrates, and the populations of these grazers often

become quite large. The grazers may build up on either the blue-green algae

or on the diatoms or unicellular green algae that precede them in the

succession of organisms in the water. At times, the unicellular green algae that

flourish support large numbers of waterfleas (cladocera), and these animals

may be so dense that they eliminate a blue-green alga that is introduced as an

inoculum (Watanabe et al., 1955a). Ostracods may sometimes be very

abundant in the paddy field, and numbers of 5,000 to 15,000 per square meter

have been recorded. An active population of such ostracods could potentially

consume 12 tons of algae per crop of rice (Grant and Alexander, 1981). These

microcrustaceans are often common in rice fields of Southeast Asia (Fernando, 1977). Grazing by Cypris sp. can prevent the development of or

suppress inocula of blue-green algae tested under laboratory conditions, and

these effects are evident in terms of the biomass or nitrogen-fixing activity of

the algae. The extent of grazing and the influence on nitrogen fixation vary

with the alga available to Cypris sp., the ostracod exhibiting a marked

preference in its feeding habits (Osa-Afiana and Alexander, 1981; Wilson et

al., 1980).

Both cladocerans and ostracods are susceptible to an array of insecticides,

and grazing by these invertebrates can be markedly reduced or totally

abolished by the use of low levels of several insecticides. As early as 1967,

Raghu and MacRae reported that the application of lindane (y-isomer of

hexachlorocyclohexane) in rice fields killed indigenous Cypris, thereby

allowing the flourishing of native blue-green algae. As little as 0.1 pg/ml of

lindane totally prevented the feeding by Cypris sp. on Tolypothrix tenuis in

laboratory tests (Grant and Alexander, 1981), and 0.1 pg of parathion per

milliliter of floodwater completely abolished the feeding on Aulosira by the

same ostracod (Osa-Afiana and Alexander, 1981). Low concentrations of

methyllindane and carbofuran (2,3-dihydro-2,2-dimethyl-7-benzofuranol

carbamate) brought about the cessation of grazing by the ostracods Cyprinotus carolinensis and Heterocypris luzonensis (Grant et al., 1983a). The

cladoceran Daphnia is also easy to control, and both lindane (Maity and

Saxena, 1979) and parathion (Watanabe et al., 1955b) are effective in this

regard. Field tests in the Philippines demonstrated that Perthane [1,ldichloro-2,2-bis(p-ethylphenyl)ethane] suppressed ostracods and resulted

not only in a 10-fold increase in biomass of blue-green algae and nitrogen

fixation but an increase in rice yield (Grant et al., 1983b). Hence, a significant

factor that often appears to reduce nitrogen fixation by blue-green algae can

be overcome by the use of insecticides. The insecticides themselves, at the low

concentrations needed to suppress grazing by the invertebrates, usually have

no significant effect on algae (Roger and Kulasooriya, 1980).

Tài liệu bạn tìm kiếm đã sẵn sàng tải về


Tải bản đầy đủ ngay(0 tr)