The element nitrogen, or "azote," meaning "without life," as Antonie Lavoisier called it about 200 years ago, has proved to be anything but lifeless, since it is a component of food, poisons, fertilizers, and explosives (277). The atmosphere contains about 10¹⁵ tonnes of N₂ gas, and the nitrogen cycle involves the transformation of some 3 × 10⁹ tonnes of N₂ per year on a global basis (244). However, transformations (e.g., N₂ fixation) are not exclusively biological. Lightning probably accounts for about 10% of the world's supply of fixed nitrogen (301). The fertilizer industry also provides very important quantities of chemically fixed nitrogen. World production of fixed nitrogen from dinitrogen for chemical fertilizer accounts for about 25% of the Earth's newly fixed N₂, and biological processes account for about 60%. Globally the consumption of fertilizer-N increased from 8 to 17 kg ha¹ of agricultural land in the 15-year period from 1973 to 1988 (107). Significant growth in fertilizer-N usage has occurred in both developed and developing countries (238). The requirements for fertilizer-N are predicted to increase further in the future (306); however, with the current technology for fertilizer production and the inefficient methods employed for fertilizer application, both the economic and ecological costs of fertilizer usage will eventually become prohibitive.

For more than 100 years, biological nitrogen fixation (BNF) has commanded the attention of scientists concerned with plant mineral nutrition, and it has been exploited extensively in agricultural practice (50, 91). However, its importance as a primary source of N for agriculture has diminished in recent decades as increasing amounts of fertilizer-N have been used for the production of food and cash crops (238). However, international emphasis on environmentally sustainable development with the use of renewable resources is likely to focus attention on the potential role of BNF in supplying N for agriculture (91, 238). The expanded interest in ecology has drawn attention to the fact that BNF is ecologically benign and that its greater exploitation can reduce the use of fossil fuels and can be helpful in reforestation and in restoration of misused lands to productivity (50, 301).

Currently, the subject of BNF is of great practical importance because the use of nitrogenous fertilizers has resulted in unacceptable levels of water pollution (increasing concentrations of toxic nitrates in drinking water supplies) and the eutrophication of lakes and rivers (19, 91, 301). Further, while BNF may be tailored to the needs of the organism, fertilizer is usually applied in a few large doses, up to 50% of which may be leached (301). This not only wastes energy and money but also leads to serious pollution problems, particularly in water supplies.

Organisms that can fix nitrogen, i.e., convert the stable nitrogen gas in the atmosphere into a biologically useful form, all belong to a biological group known as prokaryotes. All organisms which reduce dinitrogen to ammonia do so with the aid of an enzyme complex, nitrogenase. The nitrogenase enzymes are irreversibly inactivated by oxygen, and the process of nitrogen fixation uses a large amount of energy (91, 244). Nitrogenase activity is usually measured by the acetylene reduction assay, which is cheap and sensitive (91, 141, 301). The ¹⁵N isotopic method, which is also used to measure N₂ fixation, is accurate but expensive.

A wide range of organisms have the ability to fix nitrogen. However, only a very small proportion of species are able to do so; about 87 species in 2 genera of archaea, 38 genera of bacteria, and 20 genera of cyanobacteria have been identified as diazotrophs or organisms that can fix nitrogen (91, 301, 361). This wide variety of diazotrophs ensures that most ecological niches will contain one or two representatives and that lost nitrogen can be replenished.

BNF is an efficient source of nitrogen (238). The total annual terrestrial inputs of N from BNF as given by Burns and Hardy (49) and Paul (235) range from 139 million to 175 million tonnes of N, with symbiotic associations growing in arable land accounting for 25 to 30% (35 million to 44 million tons of N) and permanent pasture accounting for another 30% (45 million tons of N). While the accuracy of these figures may be open to question (301), they do help illustrate the relative importance of BNF in cropping and pasture systems and the magnitude of the task necessary if BNF is to be improved to replace a proportion of the 80 to 90 million tonnes of fertilizer-N expected to be applied annually to agricultural land by the end of the decade (238, 239). Much land has been degraded worldwide, and it is time to stop the destructive uses of land and to institute a serious reversal of land degradation (50). BNF can play a key role in land remediation.

An examination of the history of BNF shows that interest generally has focused on the symbiotic system of leguminous plants and rhizobia, because these associations have the greatest quantitative impact on the nitrogen cycle. A tremendous potential for contribution of fixed nitrogen to soil ecosystems exists among the legumes (46, 238, 313). There are approximately 700 genera and about 13,000 species of legumes, only a portion of which (about 20% [301]) have been examined for nodulation and shown to have the ability to fix N₂. Estimates are that the rhizobial symbioses with the somewhat greater than 100 agriculturally important legumes contribute nearly half the annual quantity of BNF entering soil ecosystems (313). Legumes are very important both ecologically and agriculturally because they are responsible for a substantial part of the global flux of nitrogen from atmospheric N₂ to fixed forms such as ammonia, nitrate, and organic nitrogen. Whatever the true figure, legume symbioses contribute at least 70 million tonnes of N per year, approximately half deriving from the cool and warm temperature zones and the remainder deriving from the tropics (46). Increased plant protein levels and reduced depletion of soil N reserves are obvious consequences of legume N₂ fixation. Deficiency in mineral nitrogen often limits plant growth, and so symbiotic relationships have evolved between plants and a variety of nitrogen-fixing organisms (116).

Most of the attention in this review is directed toward N₂ fixation inputs by legumes because of their proven ability to fix N₂ and their contribution to integral agricultural production systems in both tropical and temperate climates (238). Successful Rhizobium-legume symbioses will definitely increase the incorporation of BNF into soil ecosystems. Rhizobium-legume symbioses are the primary source of fixed nitrogen in land-based systems (313) and can provide well over half of the biological source of fixed nitrogen (313).

Atmospheric N₂ fixed symbiotically by the association between Rhizobium species and legumes represents a renewable source of N for agriculture (239). Values estimated for various legume crops and pasture species are often impressive, commonly falling in the range of 200 to 300 kg of N ha¹ year¹ (238). Yield increases of crops planted after harvesting of legumes are often equivalent to those expected from application of 30 to 80 kg of fertilizer-N ha¹. Inputs of fixed N for alfalfa, red clover, pea, soybean, cowpea, and vetch were estimated to be about 65 to 335 kg of N ha¹ year¹ (313) or 23 to 300 kg of N ha¹ year¹ (339). However, the measured amounts of N fixed by symbiotic systems may differ according to the method used to study N₂ fixation (279). Inputs into terrestrial ecosystems of BNF from the symbiotic relationship between legumes and their rhizobia amount to at least 70 million tons of N per year (46); this enormous quantity will have to be augmented as the world's population increases and as the natural resources that supply fertilizer-N diminish. This objective will be achieved through the development of superior legume varieties, improvements in agronomic practice, and increased efficiency of the nitrogen-fixing process itself by better management of the symbiotic relationship between plants and bacteria.

The symbioses between Rhizobium or Bradyrhizobium and legumes are a cheaper and usually more effective agronomic practice for ensuring an adequate supply of N for legume-based crop and pasture production than the application of fertilizer-N. The introduction of legumes into these pastures is seen as the best strategy to improve nitrogen nutrition of the grasses. Large contributions (between 75 and 97 kg of N ha¹ in 97 days of growth) by Stylosanthes guianensis were found (333). ¹⁵N data suggested that over 30% of the N accumulated by the grass in mixed swards could be derived from nitrogen fixed by the associated legume (333). Other recent studies (199) revealed that the nitrogen contribution of Arachis hypogaea to the growth of Zea mays in intercropping systems is equivalent to the application of 96 kg of fertilizer-N ha¹ at a ratio of plant population densities of one maize plant to four groundnut plants.

Actinorhizal interactions (Frankia-nonlegume symbioses) are major contributors to nitrogen inputs in forests, wetlands, fields, and disturbed sites of temperate and tropical regions (313). These associations involve more than 160 species of angiosperms classified among six or seven orders. The contributions of fixed nitrogen to native as well as managed ecosystems by the actinorhizal symbioses are comparable to those of the more extensively studied Rhizobium-legume interactions. Typical contributions by Alnus associations are 12 to 200 kg of N ha¹ year¹, and those by Hippophae associations are 27 to 179 kg of N ha¹ year¹ (27).

The above overview clearly indicates the significance of Rhizobium-legume symbioses as a major contributors to natural or biological N₂ fixation. Therefore, the following discussion centers on the behavior of these symbioses under severe environmental conditions and also for applications in arid regions.

EFFECTS OF SEVERE CONDITIONS ON NITROGEN FIXATION

Top Previous Next References

Environmental Conditions

Several environmental conditions are limiting factors to the growth and activity of the N₂-fixing plants. A principle of limiting factors states that "the level of crop production can be no higher than that allowed by the maximum limiting factor" (46). In the Rhizobium-legume symbiosis, which is a N₂-fixing system, the process of N₂ fixation is strongly related to the physiological state of the host plant. Therefore, a competitive and persistent rhizobial strain is not expected to express its full capacity for nitrogen fixation if limiting factors (e.g., salinity, unfavorable soil pH, nutrient deficiency, mineral toxicity, temperature extremes, insufficient or excessive soil moisture, inadequate photosynthesis, plant diseases, and grazing) impose limitations on the vigor of the host legume (46, 239, 315).

Typical environmental stresses faced by the legume nodules and their symbiotic partner (Rhizobium) may include photosynthate deprivation, water stress, salinity, soil nitrate, temperature, heavy metals, and biocides (337). A given stress may also have more than one effect: e.g., salinity may act as a water stress, which affects the photosynthetic rate, or may affect nodule metabolism directly. The most problematic environments for rhizobia are marginal lands with low rainfall, extremes of temperature, acidic soils of low nutrient status, and poor water-holding capacity (44). Populations of Rhizobium and Bradyrhizobium species vary in their tolerance to major environmental factors; consequently, screening for tolerant strains has been pursued (176). Biological processes (e.g., N₂ fixation) capable of improving agricultural productivity while minimizing soil loss and ameliorating adverse edaphic conditions are essential.

Salinity is a serious threat to agriculture in arid and semiarid regions (252). Nearly 40% of the world's land surface can be categorized as having potential salinity problems (69); most of these areas are confined to the tropics and Mediterranean regions. Increases in the salinity of soils or water supplies used for irrigation result in decreased productivity of most crop plants and lead to marked changes in the growth pattern of plants (69). Increasing salt concentrations may have a detrimental effect on soil microbial populations as a result of direct toxicity as well as through osmotic stress (313). Soil infertility in arid zones is often due to the presence of large quantities of salt, and the introduction of plants capable of surviving under these conditions (salt-tolerant plants) is worth investigating (86). There is currently a need to develop highly salt-tolerant crops to recycle agricultural drainage waters, which are literally rivers of contaminated water that are generated in arid-zone irrigation districts (129). Salt tolerance in plants is a complex phenomenon that involves morphological and developmental changes as well as physiological and biochemical processes. Salinity decreases plant growth and yield, depending upon the plant species, salinity levels, and ionic composition of the salts (86).

As with most cultivated crops, the salinity response of legumes varies greatly and depends on such factors as climatic conditions, soil properties, and the stage of growth (70-72). Variability in salt tolerance among crop legumes has been reported (353, 354). Some legumes, e.g., Vicia faba, Phaseolus vulgaris, and Glycine max, are more salt tolerant than others, e.g., Pisum sativum. It has been reported that some V. faba tolerant lines sustained nitrogen fixation under saline conditions (6, 72). Other legumes, such as Prosopis (105), Acacia (367), and Medicago sativa (7), are salt tolerant, but these legume hosts are less tolerant to salt than are their rhizobia.

The legume-Rhizobium symbioses and nodule formation on legumes are more sensitive to salt or osmotic stress than are the rhizobia (98, 330, 354, 365). Salt stress inhibits the initial steps of Rhizobium-legume symbioses. Soybean root hairs showed little curling or deformation when inoculated with Bradyrhizobium japonicum in the presence of 170 mM NaCl, and nodulation was completely suppressed by 210 mM NaCl (323). Bacterial colonization and root hair curling of V. faba were reduced in the presence of 50 to 100 mM NaCl or 100 to 200 mM polyethylene glycol as osmoticum (352, 365); the proportion of root hairs containing infection threads was reduced by 30 and 52% in the presence of NaCl and polyethylene glycol, respectively. The effects of salt stress on nodulation and nitrogen fixation of legumes have been examined in several studies (6, 9, 86, 98, 159, 223, 330, 352). The reduction of N₂-fixing activity by salt stress is usually attributed to a reduction in respiration of the nodules (86, 159, 337) and a reduction in cytosolic protein production, specifically leghemoglobin, by nodules (85, 86). The depressive effect of salt stress on N₂ fixation by legumes is directly related to the salt-induced decline in dry weight and N content in the shoot (72). The salt-induced distortions in nodule structure could also be reasons for the decline in the N₂ fixation rate by legumes subject to salt stress (302, 352, 360). Reduction in photosynthetic activity might also affect N₂ fixation by legumes under salt stress (122).

Although the root nodule-colonizing bacteria of the genera Rhizobium and Bradyrhizobium are more salt tolerant than their legume hosts, they show marked variation in salt tolerance. Growth of a number of rhizobia was inhibited by 100 mM NaCl (350), while some rhizobia, e.g., Rhizobium meliloti, were tolerant to 300 to 700 mM NaCl (99, 146, 214, 272). Strains of Rhizobium leguminosarum have been reported to be tolerant to NaCl concentrations up to 350 mM NaCl in broth culture (5, 45). Soybean and chickpea rhizobia were tolerant to 340 mM NaCl, with fast-growing strains being more tolerant than slow-growing strains (96). Rhizobium strains from Vigna unguiculata were tolerant to NaCl up to 5.5%, which is equivalent to about 450 mM NaCl (216). It has been found recently that the slow-growing peanut rhizobia are less tolerant than fast-growing rhizobia (124). Rhizobia from woody legumes also showed substantial salt tolerance: strains from Acacia, Prosopis, and Leucaena are tolerant to 500 to 850 mM NaCl (188, 364, 367). In addition to NaCl, MgCl and chlorides are more toxic than sulfates (96). It has been reported (167) that the growth of R. meliloti was severely inhibited by Mg²⁺ ions, whereas Na⁺ and K⁺ ions had little inhibitory effect.

Many species of bacteria adapt to saline conditions by the intracellular accumulation of low-molecular-weight organic solutes called osmolytes (77). The accumulation of osmolytes is thought to counteract the dehydration effect of low water activity in the medium but not to interfere with macromolecular structure or function (292). Rhizobia utilize this mechanism of osmotic adaptation (42, 43, 292, 295, 362). In the presence of high levels of salt (up to 300 to 400 mM NaCl), the levels of intracellular free glutamate and/or K⁺ were greatly increased (sometimes up to sixfold in a few minutes) in cells of R. meliloti (43, 167, 189), R. fredii (118, 119, 350), Sinorhizobium fredii (307), and rhizobia from the woody legume Leucaena leucocephala (349), K⁺ strictly controls Mg²⁺ flux during osmotic shock. An osmolyte, N-acetylglutaminyl-glutamine amide, accumulates in cells of R. meliloti (292, 294, 295); the accumulation of these osmolytes is dependent on the level of osmotic stress, the growth phase of the culture, the carbon source, and the presence of osmolytes in the growth medium.

The disaccharide trehalose plays a role in osmoregulation when rhizobia are growing under salt or osmotic stress (96, 151). Trehalose accumulates to higher levels in cells of R. leguminosarum (45) and peanut rhizobia (124) under the increasing osmotic pressure of hypersalinity. Fast-growing peanut rhizobia accumulate trehalose in the presence of many carbon sources (mannitol, sucrose, or lactose), but the slow growers accumulate trehalose only when cultured with mannitol as the carbon source. In a medium supplemented with 400 mM NaCl, the content of trehalose increased intracellularly throughout the logarithmic and stationary phases of growth of peanut rhizobia (123). The disaccharides sucrose and ectoine were used as osmoprotectants for Sinorhizobium meliloti (132). However, these compounds, unlike other bacterial osmoprotectants, do not accumulate as cytosolic osmolytes in salt-stressed S. meliloti cells.

One salt or osmotic stress response already identified in rhizobia is the intracellular accumulation of glycine betaine (189, 272, 293). The concentration of glycine betaine increases more in the salt-tolerant strains of R. meliloti than in sensitive strains (189, 293). The addition of sodium salts to bacteroids of Medicago sativa nodules increased the uptake activity of the exogenously added glycine betaine (113). These osmoprotective substances may play a significant role in the maintenance of nitrogenase activity in bacteroids under salt stress. When externally provided, glycine betaine and choline enhance the growth of Rhizobium tropici, S. meliloti, S. fredii, R. galegae, Mesorhizobium loti, M. huakuii, and Agrobacterium tumefaciens (40). However, the main physiological role of glycine betaine in the family Rhizobiaceae seems to be as an energy source, while its contribution to osmoprotection is restricted to certain strains. Another osmoprotectant, ectoine, was as effective as glycine betaine in improving the growth of R. meliloti under adverse (0.5 M NaCl) osmotic conditions (308). Ectoine does not accumulate intracellularly and therefore would not repress the synthesis of endogenous compatible solutes such as glutamate and trehalose; it may play a key role in triggering the synthesis of endogenous osmolytes (308). Therefore, at least two distinct classes of osmoprotectants exist: those such as glycine betaine or glutamate, which act as genuine osmolytes, and those such as ectoine, which act as chemical mediators.

The content of polyamines, e.g., homospermidine, increases in salt-tolerant cells and acid-tolerant strains of R. fredii (118). This polyamine may function to maintain the intracellular pH and repair the ionic imbalance caused by osmotic stress. Osmotic stress (shock) results in the formation of specific proteins in bacteria. Botsford (42) reported that the production of 41 proteins was increased at least 10-fold in salt-stressed cells of Escherichia coli. The formation of osmotic shock proteins was only recently found in cells of rhizobia. Zahran et al. (364) reported the appearance of new protein bands in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) profiles of rhizobia from woody legumes grown under salt stress. The synthesis pattern of proteins and amino acids (free or total) changes in cowpea rhizobia after high-salt (10% NaCl = 1.64 M NaCl) stress treatment (362, 365a). The Na⁺, K⁺, and Mg²⁺ concentrations are increased in cells of cowpea Rhizobium under salt stress. These organic osmolytes (amino acids) and the inorganic minerals (cations) may play a role in osmoregulation for this Rhizobium strain. Zahran et al. (362) extended their work on this halotolerant strain of cowpea Rhizobium and examined its cell morphology and ultrastructure under salt stress (1.64 M NaCl). The rhizobial cells responded to high-salt stress by changing their morphology: the cells appeared as spiral or filament-like structures, and the cell size greatly expanded. The cell ultrastructure was severely affected, the cell envelope was distorted, and the homogeneous cytoplasm was disrupted. It has been reported (51) that cells of a strain of R. meliloti appeared with irregular morphology at potentials below 0.5 MPa. Strains of rhizobia from different species modified their morphology under salt stress, and rhizobia with altered morphology have been isolated from salt-affected soils in Egypt (363). High osmotic stress (0.2 to 1.44 MPa) modified the synthesis pattern of extracellular and capsular polysaccharides of R. leguminosarum bv. trifolii (45). The colonies of R. meliloti EFB1 grown in the presence of 0.3 M NaCl show a decrease in mucoidy, and in salt-supplemented liquid medium this organism produces a 40% lower level of exopolysaccharides (193). The synthesis pattern in SDS-PAGE of lipopolysaccharides (LPS) from various species of rhizobia from cultivated legumes (355) and from woody legumes (364) was modified by salt, in the presence of which the length of side chains increased. Changing the surface antigenic polysaccharide and LPS, by salt stress, might impair the Rhizobium-legume interaction. LPS are very important for the development of root nodules (38, 312).

Successful Rhizobium-legume symbioses under salt stress require the selection of salt-tolerant rhizobia from those indigenous to saline soils (354). Rhizobium strains isolated from salt-affected soils in Egypt failed to nodulate their legume host under saline and nonsaline conditions (359a). These rhizobia showed alterations in their protein and LPS patterns (355). The genetic structure of these bacteria may also be changed (356) since they showed little DNA-DNA hybridization to reference rhizobia. The Rhizobium strains that are best able to form effective symbiosis with their host legumes at high salinity levels are not necessarily derived from saline soils (305). Graham (133) reported that salt-tolerant strains of rhizobia represent only a small percentage of all strains isolated and identified; therefore, further research in selecting salt-tolerant and effective strains of rhizobia is strongly recommended. In fact, and as indicated in recent reports, some strains of salt-tolerant rhizobia are able to establish effective symbiosis, while others formed ineffective symbiosis. Isolates of R. leguminosarum from the lentil-growing regions of the Southern Nile Valley of Egypt were salt tolerant but were not effective in N₂ fixation (212). Mutant strains of R. leguminosarum bv. viciae, which grow at 200 mM NaCl, formed ineffective nodules on roots of V. faba. These nodules failed to express nitrogenase activity (63). Some strains of Rhizobium tolerated extremely high levels of salt (up to 1.88 M NaCl) but showed significantly decreased symbiotic efficiency under salt stress (223).

Inoculation of legumes by salt-tolerant strains of R. leguminosarum bv. trifolii and R. meliloti enhanced nodulation and N content under salt stress up to 1% NaCl (95). Salt-tolerant strains isolated from Acacia redolens, growing in saline areas of Australia, produced effective nodules on both A. redolens and A. cyclops grown in sand at salinity levels up to 80 mM NaCl (75). The growth, nodulation, and N₂ fixation (N content) of Acacia ampliceps, inoculated with salt-tolerant Rhizobium strains in sand culture, were resistant to salt levels up to 200 mM NaCl (370). Under saline conditions, the salt-tolerant strains of Rhizobium sp. formed more effective N₂-fixing symbiosis with soybean than did the salt-sensitive strains (97). An important result was obtained from the recent work of Lal and Khanna (188), who showed that the rhizobia isolated from Acacia nilotica in different agroclimatic zones, which were tolerant to 850 mM NaCl, formed effective N₂-fixing nodules on Acacia trees grown at 150 mM NaCl. It was concluded from these results that salt-tolerant strains of Rhizobium can nodulate legumes and form effective N₂-fixing symbioses in soils with moderate salinity. Therefore, inoculation of various legumes with salt-tolerant strains of rhizobia will improve N₂ fixation in saline environments (370). However, tolerance of the legume host to salt is the most important factor in determining the success of compatible Rhizobium strains to form successful symbiosis under conditions of high soil salinity (75). Evidence presented in the literature suggests a need to select plant genotypes that are tolerant to salt stress and then match them with the salt-tolerant and effective strain of rhizobia (70, 329). In fact, the best results for symbiotic N₂ fixation under salt stress are obtained if both symbiotic partners and all the different steps in their interaction (nodule formation, activity, etc.) resist such stress (34, 122, 364).

The use of actinorhizal associations to improve N₂ fixation in saline environments was also studied but not as extensively as Rhizobium-legume associations. One of these actinorhizal associations (Frankia-Casuarina) is known to operate in dry climates and saline lands and was reported to be tolerant to salt up to 250 to 500 mM NaCl (67, 94). Casuarina obesa plants are highly salt tolerant (254), but growth under saline conditions depends on the effectiveness of symbiotic N₂ fixation. Successful plantings of Casuarina in saline environments require the selection of salt-tolerant Frankia strains to form effective N₂-fixing association.

The occurrence of rhizobial populations in desert soils and the effective nodulation of legumes growing therein (164, 165, 336) emphasize the fact that rhizobia can exist in soils with limiting moisture levels; however, population densities tend to be lowest under the most desiccated conditions and to increase as the moisture stress is relieved (313). It is well known that some free-living rhizobia (saprophytic) are capable of survival under drought stress or low water potential (117). A strain of Prosopis (mesquite) rhizobia isolated from the desert soil survived in desert soil for 1 month, whereas a commercial strain was unable to survive under these conditions (284). The survival of a strain of Bradyrhizobium from Cajanus in a sandy loam soil was very poor; this strain did not persist to the next cropping season, when the moisture content was about 2.0 to 15.5%. The survival and activity of microorganisms may depend on their distribution among microhabitats and changes in soil moisture (231). The distribution of R. leguminosarum in a loamy sand and silt loam soil was influenced by the initial moisture content (245). Moderate moisture tension slowed the movement of R. trifolii (139); the migration of bacteria ceased when water-filled pores in soil became discontinuous as a result of water stress. The migration of strains of B. japonicum from the initial point of placement was found to be very limited (335); the effective strains migrated into the soil to a greater extent than the ineffective strains did.

One of the immediate responses of rhizobia to water stress (low water potential) concerns the morphological changes. Mesquite Rhizobium (284) and R. meliloti (51) showed irregular morphology at low water potential. The modification of rhizobial cells by water stress will eventually lead to a reduction in infection and nodulation of legumes. Low water content in soil was suggested to be involved in the lack of success of soybean inoculation in soils with a high indigenous population of R. japonicum (156). Further, a reduction in the soil moisture from 5.5 to 3.5% significantly decreased the number of infection threads formed inside root hairs and completely inhibited the nodulation of T. subterraneum (345). Similarly, water deficit, simulated with polyethylene glycol, significantly reduced infection thread formation and nodulation of Vicia faba plants (352, 365). A favorable rhizosphere environment is vital to legume-Rhizobium interaction; however, the magnitude of the stress effects and the rate of inhibition of the symbiosis usually depend on the phase of growth and development, as well as the severity of the stress. For example, mild water stress reduces only the number of nodules formed on roots of soybean, while moderate and severe water stress reduces both the number and size of nodules (342).

Symbiotic N₂ fixation of legumes is also highly sensitive to soil water deficiency. A number of temperate and tropical legumes, e.g., Medicago sativa (7, 21), Pisum sativum (5), Arachis hypogaea (286), Vicia faba (5, 138, 365), Glycine max (89, 179, 251), Vigna sp. (233, 331), Aeschynomene (15), and the shrub legume Adenocarpus decorticand (215) exhibit a reduction in nitrogen fixation when subject to soil moisture deficit. Soil moisture deficiency has a pronounced effect on N₂ fixation because nodule initiation, growth, and activity are all more sensitive to water stress than are general root and shoot metabolism (14, 365). The response of nodulation and N₂ fixation to water stress depends on the growth stage of the plants. It was found that water stress imposed during vegetative growth was more detrimental to nodulation and nitrogen fixation than that imposed during the reproduction stage (236). There was little chance for recovery from water stress in the reproductive stage. Nodule P concentrations and P use efficiency declined linearly with soil and root water content during the harvest period of soybean-Bradyrhizobium symbiosis (115). More recently, Sellstedt et al. (279) found that N derived from N₂ fixation was decreased by about 26% as a result of water deficiency when measured by the acetylene reduction assay.

The wide range of moisture levels characteristic of ecosystems where legumes have been shown to fix nitrogen suggests that rhizobial strains with different sensitivity to soil moisture can be selected. Laboratory studies have shown that sensitivity to moisture stress varies for a variety of rhizobial strains, e.g., R. leguminosarum bv. trifolii (117), R. meliloti (51), cowpea rhizobia (232), and B. japonicum (196). Thus, we can reasonably assume that rhizobial strains can be selected with moisture stress tolerance within the range of their legume host. Optimization of soil moisture for growth of the host plant, which is generally more sensitive to moisture stress than bacteria, results in maximal development of fixed-nitrogen inputs into the soil system by the Rhizobium-legume symbiosis (313).

Drought-tolerant, N₂-fixing legumes can be selected, although the majority of legumes are sensitive to drought stress. Moisture stress had little or no effect on N₂ fixation by some forage crop legumes, e.g., M. sativa (175), grain legumes, e.g., groundnut (Arachis hypogaea) (331), and some tropical legumes, e.g., Desmodium intortum (13). One legume, guar (Cyamopsis tetragonoloba), is drought tolerant and is known to be adapted to the conditions prevailing in arid regions (332). Variability in nitrogen fixation under drought stress was found among genotypes of Vigna radiata (248) and Trifolium repens (262). These results assume a significant role of N₂-fixing Rhizobium-legume symbioses in the improvement of soil fertility in arid and semiarid habitats.

Several mechanisms have been suggested to explain the varied physiological responses of several legumes to water stress. The legumes with a high tolerance to water stress usually exhibit osmotic adjustment; this adjustment is partly accounted for by changing cell turgor and by accumulation of some osmotically active solutes (112). The accumulation of specific organic solutes (osmotica) is a characteristic response of plants subject to prolonged severe water stress. One of these solutes is proline, which accumulates in different legumes, e.g., Glycine max (120) and Phaseolus vulgaris (172). In these plants, positive correlations were found between proline accumulation and drought tolerance. Other compounds, e.g., the free amino acids and low-molecular-weight solutes such as pinitol (o-methylinositol), accumulate in several tropical legumes under drought stress (112, 185). Potassium is known to improve the resistance of plants to environmental stress. A recent report (269) indicates that K can apparently alleviate the effects of water shortage on symbiotic N₂ fixation of V. faba and P. vulgaris. The presence of 0.8 or 0.3 mM K⁺ allowed nodulation and subsequent nitrogen fixation of V. faba and P. vulgaris under a high-water regimen (field capacity to 25% depletion). It was also shown that the symbiotic system in these legumes is less tolerant to limiting K supply than are the plants themselves. Species of legumes vary in the type and quantity of the organic solutes which accumulate intracellularly in leguminous plants under water stress. This could be a criterion for selecting drought-tolerant legume-Rhizobium symbioses that are able to adapt to arid climates.

High soil temperatures in tropical and subtropical areas are a major problem for biological nitrogen fixation of legume crops (210). High root temperatures strongly affect bacterial infection and N₂ fixation in several legume species, including soybean (218), guar (22), peanut (180), cowpea (249), and beans (154, 241). Critical temperatures for N₂ fixation are 30°C for clover and pea and range between 35 and 40°C for soybean, guar, peanut, and cowpea (210). Nodule functioning in common beans (Phaseolus spp.) is optimal between 25 and 30°C and is hampered by root temperatures between 30 and 33°C (241). Nodulation and symbiotic nitrogen fixation depend on the nodulating strain in addition to the plant cultivar (22, 218). Temperature affects root hair infection, bacteroid differentiation, nodule structure, and the functioning of the legume root nodule (265, 266).

High (not extreme) soil temperatures will delay nodulation or restrict it to the subsurface region (133). Munns et al. (221) found that alfalfa plants grown in desert environments in California maintained few nodules in the top 5 cm of soil but were extensively nodulated below this depth. Nodulation of soybean was markedly inhibited at 42 and 45°C during 12-h and 9-h days, respectively (186), with no correlation between the ability of plant strains to grow at high temperature and to induce nodulation under temperature stress. Piha and Munns (241) reported that bean nodules formed at 35°C were small and had low specific nitrogenase activity, and Hernandez-Armenta et al. (148) found that transferring nodulated bean plants from a daily temperature of 26 to 35°C markedly inhibited nitrogen fixation. Some soybean varieties appear somewhat more heat tolerant, with nitrogen fixation being severely inhibited only by daytime temperatures higher than 41°C (187). The acetylene reduction activity of nodulated roots excised from unstressed bean plants (Phaseolus) was strongly diminished at 35 or 40°C when plants were nodulated by heat-sensitive or heat-tolerant strains (210).

For most rhizobia, the optimum temperature range for growth in culture is 28 to 31°C, and many are unable to grow at 37°C (133). However, 90% of cowpea Rhizobium strains obtained from the hot, dry environment of the Sahel Savannah grew well at 40°C (93). Strain adaptation to high temperature has also been reported by Hartel and Alexander (144) and Karanja and Wood (173). The latter authors found that a high percentage of the strains that persisted at 45°C lost their infectiveness. They attributed these losses in infectiveness to plasmid curing. Heat treatment of R. phaseoli at 35 and 37°C resulted in mutant strains lacking a plasmid DNA implicated in the synthesis of melanin and is related to the loss of symbiotic properties of these bacteria (36). Screening of R. leguminosarum bv. phaseoli showed that some strains were able to nodulate Phaseolus vulgaris at high temperatures (35 and 38°C) but that the nodules formed at high temperatures were ineffective and plants did not accumulate N in shoots (154).

Rhizobial survival in soil exposed to high temperature is greater in soil aggregates than in nonaggregated soil and is favored by dry rather than moist conditions (133). Ten inoculant strains of Rhizobium spp. examined by Somasegaran et al. (297) showed a gradual decline in population during 8 weeks of incubation at 37°C, while exposure to 46°C was lethal to all strains in less than 2 weeks. A decrease in the infectivity of cowpea rhizobia following storage at 35°C has also been documented (343). High soil temperature could contribute to the frequency of noninfective isolates in soil; Segovia et al. (278) noted that such noninfective isolates actually outnumbered those that were infective in the rhizosphere of bean. R. leguminosarum isolates from lentil plants in the Southern Nile Valley of Egypt were tolerant to 35 to 40°C; however, these heat-tolerant rhizobia formed less effective symbiosis with their legume hosts (212). Several heat-tolerant N₂-fixing bean-nodulating Rhizobium strains (which grow at 40°C) have been described recently (155, 210).

Heat shock proteins have been found in Rhizobium (1) but have not been studied in detail (133). The synthesis of heat shock proteins was detected in both heat-tolerant and heat-sensitive bean-nodulating Rhizobium strains (210) at different temperatures. An increased synthesis of 14 heat shock proteins in heat-sensitive strains and of 6 heat shock proteins in heat-tolerant strains was observed at 40 and 45°C, respectively (210). Heat-tolerant rhizobia are likely to be found in environments affected by temperature stress. Rhizobia isolated from the root nodules of Acacia senegal and Prosopis chilensis, growing in hot, dry regions of Sudan, had high maximum growth temperatures (44.2°C) (364, 367). Heat stress (35 and 40°C) changed the pattern of LPS mobility of some strains of tree rhizobia, as shown by SDS-PAGE (364). The same authors found that temperature stress consistently promoted the production of a protein with a relative mobility of 65 kDa in four strains of tree legume rhizobia. The 65-kDa protein that was detected under heat stress was heavily overproduced. This protein was not overproduced during salt or osmotic stress (364), which indicates that it is a specific response to heat stress.

Soil acidity is a significant problem facing agricultural production in many areas of the world and limits legume productivity (41, 65, 73, 133). Most leguminous plants require a neutral or slightly acidic soil for growth, especially when they depend on symbiotic N₂ fixation (41, 47). It has been recently reported (207, 309) that pasture and grain legumes acidify soil to a greater extent and that the legume species differ in their capacity to produce acids. Legumes and their rhizobia exhibit varied responses to acidity. Some species, like lucerne (M. sativa), are extremely sensitive to acidity, while others, such as Lotus tenuis, tolerate relatively low soil pH (73). Soil acidity constrains symbiotic N₂ fixation in both tropical and temperate soils (220), limiting Rhizobium survival and persistence in soils and reducing nodulation (47, 136, 157). Rhizobia with a higher tolerance to acidity have been identified (136). These strains usually but not always perform better under acidic soil conditions in the field (134). It has been found that R. loti multiplied at pH 4.5 but Bradyrhizobium strains failed to multiply (68); the acid-tolerant strains of R. loti demonstrate a comparative advantage over acid-sensitive strains in the ability to nodulate their host legume at pH 4.5. R. tropici and R. loti are moderately acid tolerant (344), while R. meliloti is very sensitive to acid stress (47, 318). However, R. meliloti WSM 419 has recently been shown (65) to perform satisfactorily in the field in acidic soils (pH 5.0 to 5.5). Strains of a given species vary widely in certain cases in their pH tolerance. The fast-growing strains of rhizobia have generally been considered less tolerant to acid pH than have slowly growing strains of Bradyrhizobium (134), although some strains of the fast-growing rhizobia, e.g., R. loti and R. tropici, are highly acid tolerant (68, 78, 134, 344). Recent reports, however, support the existence of acid-tolerant fast-growing strains, since both fast- and slow-growing strains of Vigna unguiculata which are tolerant to pH values as low as 4.0 have been isolated (216). The basis for differences in pH tolerance among strains of Rhizobium and Bradyrhizobium is still not clear (73, 134), although several workers have shown that the cytoplasmic pH of acid-tolerant strains is less strongly affected by external acidity (60, 62, 131, 230). Aarons and Graham (1) reported high cytoplasmic potassium and glutamate levels in acid-stressed cells of R. leguminosarum bv. phaseoli, a response which is similar to that found in osmotically stressed cells. Differences in LPS composition, proton exclusion and extrusion (60, 62), accumulation of cellular polyamines (118), and synthesis of acid shock proteins (150) have been associated with the growth of cells at acid pH. The composition and structure of the outer membrane could also be a factor in pH tolerance (134). Studies on the genetic basis of tolerance to low pH suggest that at least two loci of either megaplasmid or chromosomal location for pH genes are necessary for the growth of rhizobia at low pH (60-62). Acid tolerance in R. loti (73) was related to the composition and structure of the membrane, and acid-tolerant strains showed one membrane protein of 49.5 kDa and three soluble proteins of 66.0, 85.0, and 44.0 kDa. The expression of these proteins increased when the cells were grown at pH 4.0. The same authors (73) suggested that acid tolerance in R. loti involves constitutive mechanisms, such as permeability of the outer membrane, together with adaptive responses, including the state of bacterial growth and concomitant changes in protein expression.

The failure of legumes to nodulate under acid-soil conditions is common, especially in soils of pH less than 5.0. The inability of some rhizobia to persist under such conditions is one cause of nodulation failure (30, 55, 136), but poor nodulation can occur even where a viable Rhizobium population can be demonstrated (133, 134). Evans et al. (104) found that nodulation of P. sativum was 10 times more susceptible to acidity than was either rhizobial multiplication or plant growth. Some legumes, e.g., Trifolium subterranean, T. balansae, Medicago murex, and M. truncatula, showed tolerance to soil acidity as indicated by dry-matter yield; however, the establishment of nodules was more sensitive to soil acidity in most of these plants than was indicated by the relative yields of dry matter (102). Despite this, elevated inoculation levels have enhanced the nodulation response under acidic conditions in some studies (243). The growth, nodulation, and yield of V. faba were improved after inoculation with strains of R. leguminosarum bv. viciae in acid soils (55). It appears that the pH-sensitive stage in nodulation occurs early in the infection process and that Rhizobium attachment to root hairs is one of the stages affected by acidic conditions in soils (54, 326). Taylor et al. (314) reported that acidity had more severe effects on rhizobial multiplication than did Al stress and low P conditions. They suggested that colonization of soils and soybean roots by B. japonicum may be adversely affected by acidity, an effect which will result in reduced nodulation.

The host cultivar-rhizobial strain interaction at acid pH has also been investigated. Munns et al. (221) noted that nodulation and nitrogen fixation by some strains of Bradyrhizobium at acidic pH differ with the cultivar of mung bean used. Vargas and Graham (327) examined the cultivar and pH effects on competition for nodule sites between isolates of Rhizobium in beans (P. vulgaris) under acidic conditions. They found a significant effect of host cultivar, ratio of inoculation, and pH on the percentage of nodule occupancy by each strain. However, it has been suggested (326) that only one of the symbionts needed to be acid tolerant for good nodulation to be achieved at pH 4.5. Inoculation of Medicago polymorpha by an acid-tolerant R. meliloti strain has extended the area of acidic soils in Western Australia that can be sown with annual legumes to some 350,000 ha (153). The performance of the R. trifolii-Trifolium pratense symbiosis under acidic conditions is best when the rhizobial strains were isolated from the most acidic soils, i.e., acid-tolerant strains (191). Rhizobia appear to be vary in their symbiotic efficiency under acidic conditions. Van Rossum et al. (325) compared 12 strains of Bradyrhizobium for their symbiotic performance with groundnut in acidic soils and found that some strains were totally ineffective under acidic stress (pH 5.0 to 6.5) while others performed well under these conditions. Acid-tolerant alfalfa-nodulating strains of rhizobia, isolated from acidic soils, were able to grow at pH 5.0 and formed nodules in alfalfa with a low rate of nitrogen fixation (87). The results also demonstrate the complexity of the rhizobial populations present in the acidic soils, represented by a major group of nitrogen-fixing rhizobia and a second group of ineffective and less predominant isolates.

The host legume appears to be the limiting factor for establishing Rhizobium-legume symbiosis under acidic conditions. Legume species differ greatly in their response to low pH with regard to growth and nodulation (311). Recently, it has been found that the amount of N₂ fixed by forage legumes on low-fertility acidic soil is dependent on legume growth and persistence (316). However, selection of acid-tolerant rhizobia to inoculate legume hosts under acidic conditions will ensure the establishment of the symbiosis and also successful performance (73, 128, 229).

Recent reports indicated the destructive effects of acidic soils on Rhizobium-legume symbiosis and N₂ fixation. Low pH reduced the number of R. leguminosarum bv. trifolii cells in soils, which resulted in no or ineffective nodulation by clover plants (157). The number of nodules, the nitrogenase activity, the nodule ultrastructure, and the fresh and dry weights of nodules were affected to a greater extent at a low medium pH (<4.5) (328).

In acidic soils with pH of >5.0, where heavy-metal activity is relevant, the presence of available aluminum inhibits nodulation (35, 41). Rhizobia showed varied responses to aluminum toxicity in acidic soils and cultures. Strains of Rhizobium (326, 344) and Bradyrhizobium (133) that were resistant to aluminum (50 µM) at low pH (>5.0) were identified; however, rhizobia from clover were sensitive to these conditions (344). Johnson and Wood (168) reported that Al was taken up and bound to the DNA of both sensitive and tolerant strains but that DNA synthesis by the tolerant strains of R. loti was not affected. However, Richardson et al. (261) found that 7.5 µM Al depressed nod gene expression at low pH (4.8).

Legume species vary markedly in their tolerance to Al³⁺ and Mn²⁺, with some plants being significantly more strongly affected by these ions than are the rhizobia (133). Therefore, for acid soils with high Al content, improvement is achieved by manipulating the plant rather than the rhizobia (314). Nodulation of legumes appears more sensitive to Al than does plant growth (133); at pH 4.5 and with 0.5 mM Ca²⁺, nodulation in cowpea was delayed by 12.7 µM Al and nodule number and dry weight were severely depressed (20). Availability of Ca²⁺ in acidic soils with a high Al content appears very important for nodulation; a low Ca²⁺ concentration (0.13 mM) at pH 4.5 greatly affected nodule number, nitrogenase activity, and nodule ultrastructure of the common bean, Phaseolus (328).

Two strategies have been adopted to solve the problem of soil acidity: (i) selecting tolerant plants, and (ii) liming the acidic soil to ameliorate the effects of acidic conditions. Few cultivated legumes are adapted to low pH levels. The primary protective mechanism of acid tolerance in certain cultivars of lentil (Lens culinaris) is excess production of citric, malic, aspartic, glucenic, and succinic acids in root exudate under acidic conditions (247). It has been recently reported that some pasture legumes acidified soils to a greater extent and that the amount of acid produced per gram of shoot dry matter (specific acid production) varied between species and with growth stages, ranging from 44 to 128 cmol/kg of dry shoot matter (309). Similarly, some grain legumes produced large amounts of acids (207), with the production of H⁺ ranging between 77 and 136 cmol/kg of dry matter. It has been suggested that Al-tolerant (acid-tolerant) plant species contain and exude more organic acid and other ligands that form stable chelates with Al and thereby reduce its chemical activity and toxicity (114).

In recent studies, trials were performed to study the effects of treating soil acidity by applying lime (at rate of 2,500 kg ha¹) and superphosphate (at rates up to 20 kg ha¹) to acidic soils (239). The amelioration increased the soil pH from 4.5 to 4.9, decreased the concentration of extractable Al and Mn, and improved growth and N₂ fixation of T. subterranean. Amelioration of subsoil acidity was also done by applying coal-derived calcium fulvate (324), and this treatment increased the pH more than did amelioration by gypsum, Ca-EDTA, Ca (OH)₂, or CaCO₃. Previous reports also indicate the importance of liming for improvement of growth and nodulation of legumes in acidic soils, since they indicated that liming raised the pH from 5 to 6.5 and increased the percentage of nodule occupancy of T. subterraneum (17). However, amelioration by lime and other substances, e.g., carbonate, must be optimized to avoid increasing the pH to a level which would be inhibitory to growth and symbiotic performance of legumes. Applied carbonate was found to react with Na and raise the pH (41). Addition of bicarbonate decreased nodulation, growth, and shoot nitrogen in some grain legumes (311). Nodulation inhibition in Lupinus angustifolius grown in a limed sand at a pH of >7.0 has also been reported. Nodulation of groundnut (Arachis hypogaea) was also inhibited when plants grew in nutrient solution containing carbonate (309). High pH (>6.0 and up to 10.0) totally inhibited the nodulation of some lupins (310), and the authors suggested that pH values above 6.0 have a specific effect in the impairment of nodulation of lupins. However, rhizobia appear to be more tolerant to alkalinity than do their legume hosts. The number of R. leguminosarum bv. trifolii cells was greater in carbonate-treated soil (103); increasing the soil pH increased both the rate at which rhizobia colonized the soil and the frequency of nodule formation. It has also been reported (253) that while germination of pigeon pea was decreased at pH values of >8.8, growth of rhizobia was unaffected up to pH 11.5. These authors also found that uninoculated pigeon pea plants had as good a nodulation as did those grown from plant seeds inoculated with Rhizobium in reclaimed alkaline soils in a greenhouse study.

The tolerance of actinorhizal plants to soil acidity and acidic conditions was also reported. Solution culture studies have shown reduced nodulation of black alder (Alnus glutinosa) and other actinorhizal plants at low pH. The effect of soil acidity on nodulation of A. glutinosa grown in mine soils, limed to various pH values, was also studied (137). The authors found that soil pH was a significant factor affecting nodulation in the mine soil, with the highest level of nodulation occurring between soil pH values of 5.5 and 7.2 and the level being reduced below pH 5.5. There was also evidence of decreased viability of the endophyte (Frankia) below pH 4.5 (137). In a recent study, Igual et al. (158) reported a decrease in nodulation of Casuarina cunninghamiana at high levels of Al, with the nitrogen-fixing efficiency being decreased from 0.20 to 0.10 mg of N fixed per mg (dry weight) of nodule at 880 µM Al³⁺. They found that the mean N concentration of nodules was significantly lower at pH 4.0 (1.83%) than at pH 6.0 (2.01%).

Soil salinity and acidity are usually accompanied by mineral toxicity (specific ion toxicity), nutrient deficiency, and nutrient disorder. Salt damage to nonhalophytic plants grown in nutrient solution is often due to the effect of ion imbalance (disorder) rather than the osmotic potential (347). This disorder might occur by specific toxicity of ions such as Na⁺ and Cl and might be balanced by increasing the concentration of counterions, like K⁺ and Ca²⁺, against Cl (127). It has been suggested that K⁺ and NO₃ inhibited Na⁺ and Cl translocation from the roots to the shoots of Arachis hypogaea, so that leaf growth was protected against salt damage (285). The dominant ions in saline waters and saline soils which are available in arid zones are Na⁺ and Cl. Excess Na⁺ often harms nonhalophytes by displacing Ca²⁺ from root membranes and thus changing their integrity and their normal functioning (76). Also, acidic stress markedly affects ion absorption by and growth of roots (320); the membrane structure and function of the roots suffer fatal changes under these stress conditions. The requirement of some essential elements, e.g., Ca²⁺ and P, is increased under severe stress conditions. The requirement of Ca²⁺ for growth of R. meliloti was increased under osmotic stress (51). The Ca-depleted cells of R. leguminosarum are swollen, lack rigidity, and express an additional somatic antigen normally blocked by side chains of the LPS O antigen (88). High levels of salinity (up to 10% NaCl) decreased the Ca²⁺ content of Rhizobium cells (362), and the outer membrane structure of the Rhizobium cells was greatly distorted. In the same way, calcium appears significantly more important in cells exposed to low pH (133). O'Hara et al. (230) found that in acid-sensitive strains of R. meliloti, 1.2 mM Ca²⁺ was needed for cytoplasmic pH maintenance, and Beck and Munns (32) noted that phosphorus-limited cells or cells grown at low pH needed Ca²⁺ for phosphorus mobilization in the cell. Lack of Ca²⁺ produced some changes in ion transport, which are caused mostly by changes in membrane properties (66), and Ca²⁺ plays an essential role in cell division, elongation, and membrane structure and function (320). At low pH, addition of Ca²⁺ to the incubation medium improves both growth and ion uptake by roots (320); it was also suggested that Ca²⁺ offset the harmful effects of ions such as K⁺ and H⁺ and of stress. Ca²⁺ seems to have two effects on K transport, (i) control of K⁺ permeability and (ii) activation of K⁺ uptake through the acidification of the cytoplasm. Calcium-dependent cell surface components affect the attachment of Rhizobium to root hair cells (54, 293). It has been found (365) that salt stress (100 mM NaCl) reduced the attachment to and colonization of root hairs of V. faba plants by R. leguminosarum; this was attributed to the effect of salt on calcium availability. The effects of salt stress or acidity on calcium availability and on the initial stages of nodule formation will affect the net nodulating capacity of legumes. Both pH (4.5) and aluminum (100 µM) caused delays in nodulation of Vigna unguiculata, particularly at low Ca²⁺ levels (0.3 mM), while at a high calcium concentration (3.0 mM), nodulation was improved (152). The critical Ca²⁺ level for nodule formation in pigeon pea and guar is more than 50 µM, whereas peanut and cowpea nodulated very well in solution culture with 2 µM calcium (35). Nodulation and nodule development in cowpea were strongly depressed at low pH (4.5 to 5.5) and low calcium concentration (0.05 to 2.5 mM) (20). Stress conditions may inhibit nodulation of legumes through the inhibition of genetic activity. It has been reported that Ca²⁺ (10 mM) increased nod gene induction and expression activities of clover plants 5- to 10-fold at pH 4.5 to 5.2 (261).

Phosphorus is one of several elements which affects N₂ fixation, and, along with N, it is a principal yield-limiting nutrient in many regions of the world (240). Strains of rhizobia differ markedly in tolerance to phosphorus deficiency (33). Rhizobial P deficiency when there is a P deficiency in the soil and rhizosphere is a real possibility, especially under acidic conditions, where dissolved phosphorus salts may be precipitated in the presence of aluminum (133). Slow-growing strains of rhizobia appear more tolerant to low P levels than do fast-growing R. meliloti, in particular (33); this bacterium failed to grow at 0.06 µM P, regardless of the Ca²⁺ concentration, and some strains needed high Ca²⁺ levels to grow at 0.5 and 5.0 µM P. Phosphate-limited cultures of both fast- and slow-growing rhizobia do take up phosphate 10- to 180-fold faster than cells grown with adequate P (291), and inducible alkaline phosphatase activity was detected in P-limited cells of fast-growing R. trifolii strains (290, 291). Recently, it has been reported (18) that free-living R. tropici and bacteroids respond to P stress by increasing their P transport capacity and inducing both acid and alkali phosphatases. This P stress response occurred when the medium P concentration decreased below 1 µM. Leguminous species differ in their phosphorus requirement for growth from 0.8 to 3 µM (110).

Phosphorus appears essential for both nodulation and N₂ fixation (240, 303). Nodules are strong sinks for P and range in P content from 0.72 to 1.2% (142, 143); as a consequence, N₂ fixation-dependent plants will require more of this element than those supplied with combined nitrogen (56, 57). Nodulation, N₂ fixation, and specific nodule activity are directly related to the P supply (163, 190, 288). Application of KH₂PO₄ (25 mg of P per kg of soil) to acidic soils significantly increased the percent nodule occupancy of Trifolium subterranean by R. leguminosarum bv. trifolii (17). The nodulation and N₂ fixation (nitrogenase activity) of T. vesiculosum increased significantly after the addition of P (100 mg per kg of soil) and K (300 mg per kg of soil); nitrogenase activity was doubled when the P concentration increased to 400 mg per kg of soil (194). The interaction of P and Zn and their effects on nodulation of legumes under salt stress were studied. Saxena and Rewari (273) found that application of phosphate (20 and 40 ppm) improved the growth and nodulation of chickpea (C. arietinum) in the presence of Zn²⁺ (5 ppm) at two levels (4.34 and 8.3 dS/m) of salinity. They suggested that augmentation with Zn²⁺ provided protection to the plant under saline conditions by reducing the Na⁺/K⁺ ratio in the shoot; the shoot N content after augmentation with Zn²⁺ and in the presence of phosphate was equal to that of nonsaline control. Differences between cultivars of some legume species with regard to phosphorus requirements have been reported (48). Variability of N₂ fixation under low P availability existed between lines of P. vulgaris; high N₂-fixing and high-yielding progeny lines were detected (240).

Mycorrhizal infection of roots of legumes has been reported to stimulate both nodulation and N₂ fixation, especially in soils low in available P (121, 257). The role of mycorrhizal fungi in the protection of the Medicago sativa-R. meliloti symbiosis against salt stress was recently studied (26), and it was found that the interaction between soluble P in the soil mycorrhizal inoculum and the degree of salinity in relation to concentration and nodule formation increased with the amount of plant-available P or mycorrhizal inoculum in the soil and generally declined as the salinity in the solution culture increased. Azcon and Elatrash (26) found that the mycorrhizal inoculation protected plants from salt stress more efficiently than did any amount of plant-available P in the soil, particularly at the highest salinity level applied (43.5 dS/m).

Nitrogen fixation by the Frankia-actinorhizal symbiosis may be limited by low available P in soils. Sangina et al. (270) observed increased N₂ fixation by Casuarina equisetifolia by adding phosphate to P-deficient soil, and Reddell et al. (255) found a greater increase in the yield (wood volume) of Frankia-inoculated Casuarina cunninghamiana by adding phosphate to soil. Low P status is a frequent limitation to nodulation of actinorhizal plants. It has been reported that symbiotic N₂ fixation of the Frankia-Casuarina association requires higher P levels than those required for plant growth, when the P concentration in soil is low (270). Genetic variations among species of Allocasuarina in relation to P requirement were identified; species showed different nodulation abilities in soils with low available P (270).

Sewage sludge treatment and organic fertilizers. Sewage sludge application to agricultural soils is an economical way of disposal (109, 202, 228). It improves the physical characteristics of the soil (340) and increases organic matter content and essential plant nutrients, in particular N and P (109). Sewage sludge contains numerous components required for microbial growth and may increase the activity of soil microorganisms (299), including rhizobial growth (177). Contaminants associated with certain fertilizers such as sewage sludge may also negatively affect the survival of various soil microorganisms (202). Concern about the use of sewage sludge contaminated by heavy metals has increased. Heavy metals are known to persist in the soil over long periods and have ecotoxicological effects on plants and soil microorganisms (100). There is increasing evidence of adverse effects on microbial processes related to nutrient cycling in these types of soils (228).

1

1

1

1

1

Fertilizer application. Most combined N available to crop legumes is in the form of NO₃, formed by oxidation of NH₄ from residual fertilizer and mineralization of organic N. Nitrate has been recorded in soils at levels up to 20 mM or 280 ppm (31). Both NO₃ assimilation and N₂ fixation of legumes are strongly dependent on the plant cultivar, bacterial strain, ontogeny, and environmental factors (e.g., soil NO₃ concentration, carbon and water availability, and temperature). Invariably, N₂-fixing activity is confined to areas with low NO₃ availability (31). NO₃ may allow the plant to conserve its energy, since in overall terms more energy is required to fix N₂ than to utilize NO₃. It is therefore necessary that the plant be able to detect the presence and level of NO₃ in the rooting medium and to adjust its N₂ fixation accordingly. Symbiotic N₂ fixation in field legumes takes place against a changing background of mineral N availability as a consequence of mineralization, leaching, and, often, fertilizer application (81). Uptake of fertilizer N by plants depends on soil moisture and is higher in normal than in wet soils (with uptakes of 32 and 27%, respectively) as a result of different N-leaching losses (224).