Author for correspondence: Z. Rengel Tel: +618 6488 2557 Fax: +618 6488 1050 Email: Zed.Rengel@uwa.edu.au
Crop nutrition is frequently inadequate as a result of the expansion of cropping into marginal lands, elevated crop yields placing increasing demands on soil nutrient reserves, and environmental and economic concerns about applying fertilizers. Plants exposed to nutrient deficiency activate a range of mechanisms that result in increased nutrient availability in the rhizosphere compared with the bulk soil. Plants may change their root morphology, increase the affinity of nutrient transporters in the plasma membrane and exude organic compounds (carboxylates, phenolics, carbohydrates, enzymes, etc.) and protons. Chemical changes in the rhizosphere result in altered abundance and composition of microbial communities. Nutrient-efficient genotypes are adapted to environments with low nutrient availability. Nutrient efficiency can be enhanced by targeted breeding through pyramiding efficiency mechanisms in a desirable genotype as well as by gene transfer and manipulation. Rhizosphere microorganisms influence nutrient availability; adding beneficial microorganisms may result in enhanced availability of nutrients to crops. Understanding the role of plant–microbe–soil interactions in governing nutrient availability in the rhizosphere will enhance the economic and environmental sustainability of crop production.
Plants suffer nutrient deficiency stress when the availability of soil nutrients (and/or the amount of nutrients taken up) is lower than required for sustaining metabolic processes in a particular growth stage. Deficiency may occur as a result of (i) an inherently low amount of nutrients in the soil, (ii) low mobility of nutrients in the soil, or (iii) poor solubility of given chemical forms of the nutrients.
Plant genotypes differ in the capacity to convert nonavailable forms of nutrients to available forms and to take them up. Factors underlying the differential capacities of plant genotypes to access soil nutrients include differences in the surface area of contact between roots and soil (e.g. Sadana et al., 2002) and in the composition and amount of root exudates (Rengel, 2002; Jones et al., 2004) and rhizosphere microflora (e.g. Marschner et al., 2005b), resulting in differences in the chemistry and biology of the rhizosphere.
The availability of nutrients in the rhizosphere is controlled by the combined effects of soil properties, plant characteristics, and the interaction of roots with microorganisms (Jones et al., 2004). The concentration of nutrients and their availability to plants differ between the rhizosphere and the bulk soil (e.g. Marschner et al., 2003). The standard soil chemical analyses used for determining the concentration and ‘availability’ of nutrients in the bulk soil have, at best, only indirect relevance to nutrient availability at the root surface where uptake into the root cells takes place. In addition, the fraction of soil nutrients available to various plant species and genotypes differs widely, suggesting a limited value of soil chemical analyses that attempt to determine plant-available nutrients.
This review will concentrate on differential availability of nutrients [mainly phosphorus (P) and manganese (Mn)] in the rhizospheres of different genotypes. Root morphology, exudation of organic compounds and the contribution of soil microflora to the differential capacities of genotypes to alter nutrient availability in the rhizosphere will be discussed. Other important examples of microbe–plant interactions influencing nutrient availability are covered elsewhere, for example nitrogen (N2) fixation (Peoples et al., 2004) and mycorrhizal associations (Read & Perez-Moreno, 2003).
Mechanisms governing plant capacity to alter nutrient availability in the rhizosphere
Nutrients with limited mobility in soils [P, potassium (K), iron (Fe), zinc (Zn), Mn and copper (Cu)] are transported to roots by diffusion, which is a slow process. These nutrients are usually present in relatively large total amounts in the bulk soil, but the plant-available fraction and the concentration in the soil solution in the rhizosphere may be insufficient to satisfy plant requirements (Rengel, 2001).
A number of possible mechanisms of plant adaptation to soils with low nutrient availability have been suggested (e.g. Rengel, 2001). These mechanisms can be separated into two main categories: acquisition efficiency (capacity to take up a nutrient from soils having low availability of that nutrient) and utilization efficiency (capacity to produce a large amount of organic matter per unit of nutrient taken up). We will cover only acquisition efficiency here because of its direct relevance to the nutrient management in the rhizosphere.
Plant genotypes differ in the mechanisms for acquiring nutrients from environments with low nutrient availability (Fig. 1). Nutrient-efficient genotypes may have an increased capacity (i) to exploit the soil (large root surface area), (ii) to convert nonavailable nutrient forms into plant-available forms, and/or (iii) to take up nutrients across the plasma membrane (Rengel, 2001). However, it is likely that the increased exploitation of soil volume and conversion of nonavailable nutrient forms into available forms are most important for efficient uptake, especially of nutrients with limited soil mobility.
The diffusion rate in soil increases with the concentration gradient; thus, larger amounts of nutrients are transported toward roots if a larger concentration gradient between the root surface and the bulk soil can be maintained by vigorous nutrient uptake at the root surface. However, when the capacity of root cells to take up nutrients exceeds the rate of nutrient replenishment at the root surface, the uptake rate is governed by the nutrient supply rather than by the capacity of plants to take up nutrients. Therefore, an increased capacity of root cells to take up nutrients as a result of increased expression of high-affinity nutrient uptake systems in the plasma membrane is expected to be of secondary importance as an efficiency mechanism for diffusion-supplied nutrients for which the supply to the root surface is the rate-limiting step.
Arguably, a big effect will be achieved not by genetically engineering the nutrient transporters in the plasma membrane but by transforming plants to increase exudation of chelating and other compounds into the rhizosphere (e.g. George et al., 2005; Liu et al., 2005), because these compounds convert nutrients into plant-available forms that can be transported across the root-cell plasma membrane. However, claims in at least one published report of increased exudation of carboxylates and uptake of P by transgenic plants (Herrera-Estrella, 2000) could not be confirmed (Delhaize et al., 2001). In addition, there was no positive relationship between exudation of organic compounds and P uptake in transgenic plants that were engineered to increase exudation of phytase (e.g. George et al., 2005). Hence, many problems remain to be solved before transgenic plants with altered exudation characteristics will show consistently enhanced performance in a range of nutrient deficiency environments.
Over-expression of P transporters may result in excessive accumulation of P in cells and plant death from P toxicity (cf. Smith, 2002), showing the complexity of the nutrient deficiency response. Hence, we need to understand the fine regulation of the whole sequence that comprises signal perception, transmission and response to nutrient deficiency stress before successfully transforming plants to up-regulate mechanisms responsible for increasing nutrient availability as well as the capacity to take up nutrients. Characterizing promoters specific to a nutrient deficiency response (e.g. Schunmann et al., 2004), in addition to cloning major genes involved in the plant response to nutrient deficiency, is therefore crucial for successful production of nutrient-efficient plant genotypes that modulate their response according to nutrient availability. Promoters and major genes from one plant species could be used to transform other species. At least in the case of P nutrition, this approach could be successful because the regulation of P-deficiency-induced genes appears to be conserved in different plant species, for example in Lupinus albus and Medicago sativa (Liu et al., 2005).
Rhizosphere microorganisms and nutrient availability
Compared with the soil organic matter, root exudates represent an easily degradable nutrient source for microorganisms, allowing some microbial species to proliferate rapidly in the rhizosphere. These are usually species with high growth rates and relatively high nutrient requirements such as pseudomonads (Marilley & Aragno, 1999). The amount and composition of root exudates affect the microbial community composition which in turn will influence nutrient availability.
Plants grown with deficient vs sufficient nutrient supply often have differential microbial communities in the rhizosphere (e.g. Marschner et al., 2004a). Nutrient deficiency can influence rhizosphere microorganisms either directly (by affecting their nutrition) or indirectly (by altering root morphology and exudation). L. albus produces cluster roots under P deficiency; cluster roots of different age have different microbial community compositions, as influenced by root exudation (Marschner et al., 2002). This result suggests that nutrient deficiency influences rhizosphere microorganisms indirectly (i.e. via the plant). In contrast, microbial community structure in the rhizosphere of Poaceae genotypes was correlated with P availability in the rhizosphere (Marschner et al., 2005b), suggesting a direct effect of P on rhizosphere microorganisms.
To separate direct and indirect effects of P availability on rhizosphere microorganisms, Hordeum vulgare and Cucumis sativus were grown in a P-deficient soil supplemented with P via soil or foliar application (Fig. 2; Marschner et al., 2004a). In the foliar treatment, the effect was indirect (mediated by the plant) because the rhizosphere community composition was different between P-deficient and P-supplied plants (Fig. 2).
Rhizosphere microbial communities are not always influenced by nutrient availability. P availability had no effect on microbial community composition in the rhizosphere of, for example, Sorghum bicolor (Marschner et al., 2004b).
Clearly, rhizosphere microorganisms respond to, and are influenced by, nutrient availability. In the following sections, we discuss whether such changes could play a role in the differential nutrient efficiencies of plant genotypes.
Manipulation of phosphorus availability in the rhizosphere
Although the total amount of P in the soil may be high, it is mainly (> 80%) present in forms unavailable to plants because of adsorption, precipitation, or conversion to organic forms. In acidic soils, P forms iron/aluminium (Fe/Al) phosphates and gets adsorbed to Fe/Al oxides or humic substances. In alkaline calcareous soils, P is often precipitated as calcium (Ca)-P. Organic P (mainly phytate) may represent more than 50% of total P in many soils (Osborne & Rengel, 2002).
Generally, three broad categories of P efficiency mechanisms exist in plants to increase availability and uptake of P under deficiency conditions: (i) alteration of the geometry or architecture of the root system, (ii) secretion or exudation of chemical compounds into the rhizosphere, and (iii) association with microorganisms (Fig. 3).
Plants growing in P-deficient soil allocate a greater proportion of assimilates to root growth and tend to have fine roots of a small diameter and therefore a large surface area. Fine roots, and especially root hairs (Gahoonia et al., 2001; Nigussie et al., 2003), are effective in scavenging P from the soil environment because of a large surface area of contact with the soil. One cycle of selecting Trifolium repens for longer root hairs resulted in 14% longer hairs (for references, see Rengel, 2005), indicating a potential of such a breeding approach to increase P efficiency.
Exudation of organic compounds
Under P deficiency, plants exude a wide range of organic compounds to increase mobilization of P from sparingly soluble sources (e.g. Neumann & Römheld, 1999). The genetic basis of regulation of exudation of organic compounds, now in the process of being deciphered (see Rengel, 2002; Liu et al., 2005), will provide important knowledge for breeding and transforming crop species for enhanced root exudation. However, there is little understanding of the actual reactions and processes occurring in the rhizosphere that are governed by exuded organic compounds, leading to improved P nutrition (e.g. Hinsinger, 2001; Jones et al., 2004).
Exudation of carboxylates Exuded organic acid anions may have a role both in solubilization of mineral nutrients and as growth substrates for microorganisms. Typical carboxylates (organic acid anions) found in root exudates include citrate, malate, malonate, acetate, fumarate, succinate, lactate and oxalate (see Rengel, 2002). Malate and citrate are the major root exudate components for some plants species, such as cluster-root-forming L. albus (Neumann & Römheld, 1999). Because carboxylates are excellent substrates for microbial growth, high concentrations of carboxylates may occur temporarily and only at rapidly growing root apices not yet densely colonized by microorganisms.
Carboxylates may be exuded by P-deficient roots at appreciable rates [an average rate of 0.57 nmol citrate cm−1 root h−1 for Brassica napus (Hoffland et al., 1989) or 200–400 nmol oxalate g−1 soil h−1 for Cassia spectabilis, with rhizosphere soil containing at least 29 µmol oxalate g−1 soil (Radersma & Grierson, 2004)]. Exuded carboxylates may solubilize various P complexes (see Hinsinger, 2001).
Phosphatases and phytase Plants and microorganisms increase exudation of P-hydrolysing enzymes under P deficiency. These enzymes break down organic P, thus making P available for uptake. Phosphatases are not effective in mineralizing phytate (inositol hexaphosphate), the major form of organic P in many soils. However, phytase specifically catalyses the breakdown of phytate. Roots excrete little, if any, phytase, whereas microorganisms (e.g. Aspergillus niger) exude large amounts (Richardson et al., 2001), indirectly enabling plants to utilize phytate (Osborne & Rengel, 2002). Genetic modification of plants to excrete microbial phytase (e.g. George et al., 2005) may allow plants to increase P uptake, but the effectiveness of phytase is limited by the low phytate availability in soil and binding of phytase to soil particles.
Exudation of phosphatases increases when plants are P deficient (e.g. Radersma & Grierson, 2004). When grown in an acidic P-deficient soil amended with Fe-P, the P-efficient Triticum aestivum genotype had a greater acid phosphatase activity in the rhizosphere than the inefficient genotype, with phosphatase activity correlating positively with growth and P uptake (Marschner et al., 2005b).
Changing rhizosphere pH
In calcareous soils, rhizosphere acidification by proton extrusion causes dissolution of poorly available P forms (Hinsinger et al., 2003), such as Ca-P minerals. Rhizosphere acidification is more prominent in P-efficient Phaseolus vulgaris genotypes than in P-inefficient ones, but no difference in rhizosphere acidification was noted for a range of T. aestivum and H. vulgare genotypes differing in P efficiency (see Rengel, 2005).
An effective interaction between P solubilizers (Table 1) and plants depends on (i) a high population of P solubilizers maintained in the rhizosphere over long periods, (ii) exudation of carboxylates and protons into the rhizosphere by roots and microorganisms, (iii) low P uptake by microorganisms, and (iv) positive interaction with mycorrhizal fungi or other beneficial microorganisms.
Table 1. Examples of microbial genera affecting phosphorus (P) and manganese (Mn) availability
Many microbial species have the capacity to solubilize sparingly soluble P in vitro (Table 1; see also Whitelaw, 2000). Inoculation with effective P solubilizers may increase P uptake and growth of plants (Kumar & Narula, 1999). However, it is often not clear whether the improved P uptake by plants is directly attributable to P solubilization by the introduced microbe. Combination of P solubilizers with N2 fixers or with mycorrhizal fungi is superior to inoculation with the P solubilizers alone (Kumar & Narula, 1999).
Large-scale inoculation with P solubilizers in farming practice is hampered by several factors that could diminish the effectiveness of the introduced microorganisms: (i) most soils already contain P solubilizers, so the effect of inoculation may be small; (ii) introduced strains may have poor survival in the rhizosphere because of low competitiveness against indigenous, well-adapted strains; (iii) microorganisms are selected based on their P solubilization in vitro in conditions ideal for growth and P solubilization, whereas conditions in the rhizosphere may be far from optimal, and (iv) P solubilized by the microorganisms may be unavailable to plants because microorganisms take it up. It is of utmost importance that the possible contribution of P-solubilizing microorganisms to crop P uptake be evaluated in realistic soil conditions in the field (cf. Jones et al., 2004) because the literature abounds with reports on in vitro solubilization of P that could not be repeated in field conditions (see Gyaneshwar et al., 2002).
Manipulation of manganese availability in the rhizosphere
Yield of cereals on calcareous soils is frequently limited by Mn deficiency caused by low Mn availability, rather than low Mn content in the soil (Rengel, 2000). Mn-efficient genotypes take up more Mn from soils with limited Mn availability, but the physiological mechanisms underlying Mn efficiency are poorly understood (Rengel, 2000, 2001) (Fig. 4).
As a result of root uptake and poor mobility, depletion of Mn in the rhizosphere is greater than replenishment from bulk soil, resulting in a lower concentration of Mn in the rhizosphere compared to the bulk soil. Such Mn depletion is more prominent in the rhizospheres of Mn-efficient than Mn-inefficient T. aestivum genotypes (Marschner et al., 2003). However, in another study, the genotypic difference in Mn efficiency between T. aestivum and Triticum durum genotypes was not caused by differences in chemical mobilization of Mn in the rhizosphere (Sadana et al., 2002).
The available Mn concentration was up to 2 orders of magnitude greater in the rhizospheres of three Banksia species (Banksia attenuata, Banksia ilicifolia and Banksia menziesii) than in bulk soil (Marschner et al., 2005a). An addition of 500 µg MnO2 g−1 soil before incubation doubled the available Mn concentration to 4 µg Mn g−1 soil. After 7 d of incubation, the concentration of available Mn increased more than 10-fold, indicating active populations of Mn reducers (P. Marschner, unpublished results).
Exudation of organic compounds
Medicago sativa plants exude a variety of carboxylates under Mn deficiency. The amounts of exuded citrate and malonate (and to a lesser extent fumarate, malate, oxalate and lactate) under Mn deficiency were positively correlated with the Mn efficiency of M. sativa genotypes (Gherardi & Rengel, 2003).
Manganese availability is increased in acidic rhizospheres. However, the form of N supplied, and therefore differences in rhizosphere acidification, had no effect on differential expression of Mn efficiency among H. vulgare genotypes (see Rengel, 2001) grown in calcareous soils. However, such results may be caused by two opposite effects: (i) ammonium nutrition increasing acidification in the rhizosphere and thus enhancing availability of Mn, and (ii) ammonium decreasing Mn2+ uptake as an antagonist (Husted et al., 2005). The strong pH-buffering capacity of calcareous soils may also contribute to preventing differential expression of Mn efficiency (e.g. Tong et al., 1997).
Reduction and oxidation of Mn by microorganisms (Table 1) are important components of Mn cycling in soil. Fluorescent pseudomonads are effective Mn reducers, which appear to be more abundant in the rhizospheres of some Mn-efficient T. aestivum genotypes compared with Mn-inefficient genotypes (Rengel et al., 1998).
The composition of rhizosphere bacterial communities can be assessed using the ribosomal intergenetic spacer analysis (RISA) region of the bacterial DNA. Such RISA banding patterns of the bacterial communities in the T. aestivum rhizosphere were correlated with the concentration of diethylenetriaminepentaacetic acid (DTPA)-extractable Mn in the rhizosphere, shoot dry matter and Mn content (Marschner et al., 2003), suggesting the importance of microorganisms in plant Mn uptake.
Breeding genotypes for greater nutrient efficiency
Modern agriculture has mostly been developed on the premise of supplying nutrients in sufficient, if not luxurious, quantities as synthetic fertilizers. As a consequence, crop selection and breeding have resulted in the development of cultivars highly responsive to fertilizers, but often lacking traits necessary for growth under nutrient-limiting or adverse soil conditions (Rengel, 2001, 2005). However, this tendency has been challenged by the current emphasis on developing cultivars for sustainable, low-input agriculture, fueled by increasing cost of fertilizers, government restrictions on fertilizer applications to minimize environmental damage, and the increased use of poor-quality land.
Among macronutrients, breeding for adaptation to low-nutrient environments has advanced only for P and to some extent N and K. In the case of micronutrients, some breeding efforts have been reported for Fe, Mn, Zn, Cu and boron (B) (see Rengel, 2005). Breeding such nutrient-efficient genotypes adapted to low-input agricultural ecosystems should be a matter of priority.
Growing nutrient-efficient crop genotypes on soils of low nutrient availability represents an environmentally friendly approach that would reduce land degradation by minimizing application of chemicals (i.e. fertilizers) on agricultural land (Rengel, 2001). The danger of exhaustion of soil nutrient resources (‘land mining’) would be negligible, at least for micronutrients and P (see Rengel, 2005) because the total soil supply of these nutrients is sufficient for hundreds of years of sustainable cropping by new, efficient genotypes that can gain access to the soil nutrient pools not available to inefficient genotypes.
Understanding of the role that root exudation of organic compounds plays in increasing nutrient availability in the rhizosphere is sketchy at present. The regulation of the complete exudation process and the underlying genetics need to be elucidated.
More research into the genetic basis of qualitative and quantitative differences in root exudation is required. Given that exudation of organic compounds represents a big drain of energy and resources, thorough understanding of the regulation of the whole sequence of processes culminating in exudation of organic compounds into the rhizosphere is required before practical applications become feasible. In order to have practical significance, genetic alterations would need to be such that exudation is triggered by specific environmental conditions in an appropriate dose–response manner.
Future prospects for genetically manipulating plants to enhance their capacity to alter the biology and chemistry of the rhizosphere and increase nutrient availability under given environmental conditions will be underpinned by (i) understanding the regulation of the exudation process, (ii) elucidating the regulation of root morphology, (iii) characterizing the synthesis and activity of membrane-embedded nutrient transporters, and (iv) understanding the interactions among root exudation, indigenous rhizosphere microorganisms and nutrient availability.
Bioengineering the rhizosphere by adding beneficial microorganisms will require understanding of microbe–microbe and microbe–plant interactions, enabling introduced microorganisms to show full activity in the targeted rhizosphere.
Research work in the authors’ laboratories was funded by the Australian Research Council and the Grains Research and Development Corporation.