REVIEW ARTICLE

Acquisition of phosphorus and nitrogenin the rhizosphere and plant growth promotionby microorganisms

Alan E. Richardson & Jos-Miguel Barea &Ann M. McNeill & Claire Prigent-Combaret

Received: 24 May 2008 /Accepted: 5 January 2009 /Published online: 27 February 2009# Springer Science + Business Media B.V. 2009

Abstract The rhizosphere is a complex environmentwhere roots interact with physical, chemical andbiological properties of soil. Structural and functionalcharacteristics of roots contribute to rhizosphereprocesses and both have significant influence on the

capacity of roots to acquire nutrients. Roots alsointeract extensively with soil microorganisms whichfurther impact on plant nutrition either directly, byinfluencing nutrient availability and uptake, or indi-rectly through plant (root) growth promotion. In thispaper, features of the rhizosphere that are importantfor nutrient acquisition from soil are reviewed, withspecific emphasis on the characteristics of roots thatinfluence the availability and uptake of phosphorusand nitrogen. The interaction of roots with soilmicroorganisms, in particular with mycorrhizal fungiand non-symbiotic plant growth promoting rhizobac-teria, is also considered in relation to nutrientavailability and through the mechanisms that areassociated with plant growth promotion.

Keywords Soil microorganisms . PGPR .

Mycorrhizal fungi . Exudate . Phosphorus . Nitrogen .

Uptake .Mineralization

Introduction

The rhizosphere can be defined as the zone of soilaround plant roots whereby soil properties areinfluenced by the presence and activity of the root.Changes to the physical, chemical and biologicalproperties of rhizosphere soil has significant influenceon the subsequent growth and health of plants. Interms of nutrient acquisition, both the structural and

Plant Soil (2009) 321:305339DOI 10.1007/s11104-009-9895-2

Responsible Editor: Philippe Hinsinger.

A. E. Richardson (*)CSIRO Plant Industry,PO Box 1600, Canberra 2601, Australiae-mail: alan.richardson@csiro.au

J.-M. BareaDepartamento de Microbiologa del Suelo y SistemasSimbiticos, Estacin Experimental del Zaidn, CSIC,Prof. Albareda 1,Granada 18008, Spain

A. M. McNeillUniversity of Adelaide, Soil and Land Systems,Waite Campus,Adelaide 5005, Australia

C. Prigent-CombaretUniversit de Lyon,Lyon, France

C. Prigent-CombaretUniversit Lyon 1,Villeurbanne, France

C. Prigent-CombaretCNRS, UMR 5557, Ecologie Microbienne,Villeurbanne, France

functional characteristics of roots have long beenrecognized as being important in determining thecapacity for plants to access and mediate theavailability of essential nutrients in soil and toalleviate against those that are toxic (Darrah 1993;Hinsinger 1998; Marschner 1995). Furthermore, rootsinteract with diverse populations of soil microorgan-isms which have significant implication for growthand nutrition (Curl and Truelove 1986; Bowen andRovira 1999; Mukerji et al. 2006; Brimecombe et al.2007). Soil nutrients are transferred towards the rootsurface through the rhizosphere or, in the case of rootsassociated with mycorrhizal fungi, through themycorrhizosphere, prior to acquisition.

Plants modify the physico-chemical properties andbiological composition of the rhizosphere through arange of mechanisms, which include acidificationthrough proton extrusion and the release of rootexudates. Along with changes to soil pH, root exudatesdirectly influence nutrient availability or have indirecteffects through interaction with soil microorganisms.An outstanding feature of the rhizosphere is thatrhizodeposition and root turnover account for up to40% of the carbon input into soil and clearly is the majordriver for soil microbiological processes (Grayston et al.1996; Jones et al. 2009). Interactions between plantroots and soil microorganisms are ubiquitous acrossvarious trophic levels and are an essential componentof ecosystem function. It has become increasinglyevident that root interactions with soil microorganismsare intricate and involve highly complex communitiesthat function in very heterogeneous environments(Giri et al. 2005). Microbial interactions with rootsmay involve either endophytic or free living micro-organisms and can be symbiotic, associative or casualin nature. Beneficial symbionts include N2-fixingbacteria (e.g. rhizobia) in association with legumesand interaction of roots with mycorrhizal fungi, withthe later being particularly important in relation to plantP uptake. Associative and free-living microorganismsmay also contribute to the nutrition of plants through avariety of mechanisms including direct effects onnutrient availability (e.g. N2-fixation by diazotrophsand P-mobilization by many microorganisms), en-hancement of root growth (i.e. through plant growthpromoting rhizobacteria, or PGPR), as antagonists ofroot pathogens (Raaijmakers et al. 2009) or assaphrophytes that decompose soil detritus and subse-quently increase nutrient availability through mineral-

ization and microbial turnover. Such processes arelikely to be of greater significance for nutrientavailability in the rhizosphere where there is increasedsupply of readily metabolizable carbon and wheremobilized nutrients can be more easily captured byroots.

In this review we address the acquisition of nutrientsfrom soil by plants with specific emphasis on thestructural and functional characteristics of roots thatinfluence the availability and uptake of P and N. Inparticular, the importance of soil microorganisms andtheir interactions with roots in relation to nutrientavailability is considered, along with their associatedmechanisms of plant growth promotion. The reviewcomplements previous reviews that have specificallyfocused on either plant-based traits or mechanisticprocesses associated with P and N uptake (Raghothama1999; Vance et al. 2003; Bucher 2007, Miller andCramer 2004; Jackson et al. 2008). Although therhizosphere is important for the efficient uptake of awider range of macro and micronutrients (including Fe;Lemanceau et al. 2009), the review specifically focuseson N and P which are key nutrients that limitsustainable agricultural production across much of theglobe (Tilman et al. 2002).

Mechanisms of nutrient acquisition by plant roots

Efficient capture of nutrients from soil by roots is acritical issue for plants given that in many environmentsnutrients have poor availability and may be deficient foroptimal growth. Whilst nutrient supply in soil is oftenaugmented by the application of fertilizers, the avail-ability of nutrients is governed by a wide range ofphysico-chemical parameters, environmental and sea-sonal factors and biological interactions. Competitionfor nutrient uptake across different plant species,between different roots and with microorganisms is alsosignificant (Hodge 2004). The rate of root growth andthe plasticity of root architecture along with develop-ment of the rhizosphere, through either root growth orextension of root hairs, are clearly important foreffective exploration of soil and interception ofnutrients (Lynch 1995). Biochemical changes in therhizosphere and interaction with microorganisms arealso significant. However, the importance of differentroot traits and rhizosphere-mediated processes isdependent on the nutrient in question and other factors

306 Plant Soil (2009) 321:305339

that include plant species and soil type (Tinker andNye 2000). For example, for nutrients present at lowconcentrations in soil solution and/or with poordiffusivity (e.g. P as either HPO4

2 or H2PO4, and

micronutrients, such as Fe and Zn), root growth andproliferation into new regions of soil and release ofroot exudates are of particular importance (Barber1995; Darrah 1993). In contrast, nutrients present ineither higher concentrations (e.g. K+, NH4

+), or withgreater diffusion coefficients (e.g. NO3

, SO4 and

Ca2+), are able to move more freely toward the rootthrough mass flow, where root distribution andarchitectural characteristics that facilitate water uptakeare of greater relative significance (Barber 1995;Tinker and Nye, 2000; Lynch 2005). The relativesignificance of such factors in the acquisition anduptake of P and N is therefore considered in moredetail below.

Acquisition of phosphorus by plants

Phosphorus availability and uptake

Although soils generally contain a large amount of totalP only a small proportion is immediately available forplant uptake. Plants obtain P as orthophosphate anions(predominantly as HPO4

2 and H2PO41) from the soil

solution. In most soils the concentration of orthophos-

phate in solution is low (typically 1 to 5 M; Bieleski1973) and must therefore be replenished from otherpools of soil P to satisfy plant requirements. Ortho-phosphate is rapidly depleted in the immediate vicinityof plant roots, and as such a large concentrationgradient occurs across the rhizosphere between bulksoil and the root surface (Gahoonia and Nielsen 1997;Tinker and Nye 2000; Fig. 1). However, for most soilsthe rate of diffusion of orthophosphate is insufficient toovercome localized gradients, which in most caseslimits the uptake of sufficient P. Evidence from bothmodelling and empirical studies also suggests thatactual P uptake capacity at the root surface is unlikelyto be limiting for plant growth (Barber 1995). This issupported by more recent studies on the expression ofgenes that encode for transport proteins with highaffinity for uptake of orthophosphate (e.g. Km ~3 M),which are predominantly expressed in root hair cells onthe epidermis (Mitsukawa et al. 1997; Mudge et al.2002; Schnmann et al, 2004). Whilst expression ofthese genes has been shown to facilitate the P uptakecapacity of cells in suspension culture (Mitsukawa etal. 1997), their over-expression in transgenic plants didnot result in increased P uptake by barley (Hordeumvulgare L.) when grown at a range of P concentrationsin either solution or soil (Rae et al. 2004). This isconsistent with the view that plants are well adaptedfor uptake of P from the low concentrations that are

Root

Root exudates

Sugars

Organic anions

H+ ions

Siderophores

Phosphatases

10 mM

Desorption

Solubilization

Solubilization

Mineralization

ADSORBED P (Pi)

ORGANIC P (Po)

SOIL SOLUTION P(Pi and Po)

MINERAL P (Pi)

Root hairs

Rhizosphere Soil

~0 M1 M Diffusion

Rhizospheremicroorganisms

Organic anions

H+ ions

Siderophores

Phosphatases

Mycorrhizal fungi

Mycorrhizosphere

P uptake

Fig. 1 Physiological andchemical processes that in-fluence the availability andtransformation of phospho-rus in the rhizosphere(adapted from Richardson1994; Richardson 2001)

Plant Soil (2009) 321:305339 307

typical of soil solutions as indicated by minimumuptake concentrations (Clim values) of 0.01 to 0.1 Mfor different species (Jungk 2001). On this basis it issuggested that the supply of P to the root surface, andits availability as influenced by root and microbialprocesses (as outlined below), or the capacity of rootsto exploit new regions of soil are of greater importancefor P acquisition than the kinetics associated with itsuptake.

The importance of root growth and architecture forthe efficient capture of P is well documented and inmany cases is a specific response of plants to Pdeficiency (reviewed by Hodge 2004; Lynch 2005;Raghothama 1999; Richardson et al. 2009; Vance et al.2003). Characteristics of roots that facilitate soilexploration and hence P uptake include; rapid rate ofroot elongation and high root to shoot biomass ratio(Hill et al. 2006), increased root branching and rootangle particularly in surface soils and nutrient richregions (Lynch and Brown 2001; Manske et al. 2000;Rubio et al. 2003), high specific root length (i.e. lengthper unit mass or root fineness; Silberbush and Barber,1983), the presence of root hairs (Fhse et al. 1991;Gahoonia and Nielsen 1997; Gahoonia and Nielsen2003) and, in some species, the formation of special-ized root structures such as aerenchyma (Fan et al.2003), dauciform roots in the Cyperaceae (Shane et al.2006) and proteoid roots (or cluster roots) in theProteaceae and certain Lupinus spp. (Dinkelaker et al.1995; Gardner et al. 1981; Roelof et al. 2001; Lamberset al. 2006).

Depletion of phosphorus in the rhizosphere

The extent of P depletion in both the rhizosphere andmycorrhizosphere has been highlighted in a number ofstudies. Root hairs in particular contribute to increasedroot volume and can constitute up to 70% of the totalroot surface area (Itoh and Barber 1983; Jungk 2001;Fig. 1). As such they are the major site for nutrientacquisition and can account for up to 80% of total Puptake in non-mycorrhizal plants (Fhse et al. 1991).Variation in length and density of root hairs isimportant particularly under conditions of low P andtheir contribution to P uptake has been verified throughmodelling studies and the use of root-hairless mutants(Gahoonia and Nielsen 2003; Ma et al. 2001).Mycorrhizal colonization of roots (as discussed below)similarly provides a significant increase in the effective

volume of soil explored with associated depletion ofsoil P (Tarafdar and Marschner 1994 and as modelledby Schnepf et al. 2008). Interestingly, the benefitderived from mycorrhizal fungi has been shown to beinversely associated with root hair length across a rangeof plant species, suggesting a complementary functionof these traits (Baon et al. 1994; Schweiger et al. 1995).Reduced P acquisition by a root-hairless mutant ofbarley at low soil P was similarly compensated for bythe presence of mycorrhiza (Jakobsen et al. 2005a).

Depletion of P in the rhizosphere occurs from bothinorganic and organic P which includes soluble ortho-phosphate and various forms of extractable P (bothinorganic and organic) that are widely considered to belabile. In addition, it is evident that pools of P that aremore recalcitrant to extraction (e.g. NaOH-extractableP; Fig. 2), and thus previously considered to be only ofpoor availability to plants, can also be depleted in therhizosphere. Such studies have used a range ofdifferent plants species and soil types whereby roots,root hairs and mycorrhizas are separated from bulk soilusing meshed-compartments in rhizobox systems (e.g.Chen et al. 2002; Gahoonia and Nielsen 1997; Georgeet al. 2002; Morel and Hinsinger 1999; Nuruzzaman etal. 2006; Tarafdar and Jungk 1987). Such approachesare useful in identifying different processes thatcontribute to P depletion even if they may exaggeraterhizosphere effects. For example, Chen et al. (2002)investigated P dynamics around the roots of ryegrass(Lolium perenne L.) and radiata pine (Pinus radiata D.Don) and showed a significant depletion of variouspools of P at distances of up to 2 and 5 mm from theroot surface of the two species respectively, which wasrelated to different rhizosphere properties (Fig. 2). Bothspecies also showed a significant increase in microbialbiomass around the roots and associated increases inbicarbonate-extractable organic P may be a consequenceof microbial-mediated immobilization of orthophos-phate within the rhizosphere (Richardson et al. 2005).Further work to elucidate the role of microorganisms ininfluencing P availability within the rhizosphere andthe extent to which they either complement or competewith plant processes in P acquisition is required(Jakobsen et al. 2005b).

Role of root exudates in phosphorus mobilization

The availability of P in the rhizosphere is influencedsignificantly by changes in pH and root exudates which

308 Plant Soil (2009) 321:305339

can either directly or indirectly affect nutrient availabil-ity and/or microbial activity (Fig. 1; Richardson 1994).Acidification of the rhizosphere in response to Pdeficiency has been demonstrated for a number ofspecies (see review by Hinsinger et al. 2003) and canalter the solubility of sparingly-soluble inorganic Pcompounds (particularly Ca-phosphates in alkalinesoils), or affect the kinetics of orthophosphateadsorption-desorption reactions in soil and thesubsequent availability of orthophosphate and vari-ous micronutrients in soil solution (Hinsinger andGilkes 1996; Gahoonia and Nielsen 1992; Hinsinger2001; Neumann and Rmheld 2007).

Organic anions are released into the rhizosphere inresponse to various nutritional stresses including P, Feand micronutrient deficiency and Al toxicity (seereviews by Hocking 2001; Neumann and Rmheld2007; Ryan et al. 2001). The concentration ofdifferent organic anions is typically greater in therhizosphere (around 10-fold) compared with that inbulk soil (Jones et al. 2003). Organic anions are

commonly released from roots in association withprotons which results in an acidification of therhizosphere (Dinkelaker et al. 1989; Hoffland et al.1989; Neumann and Rmheld 2002). In addition tothis change in rhizosphere pH, organic anions canalso directly facilitate the mobilization of P throughreduced sorption of P by alteration of the surfacecharacteristics of soil particles, desorption of ortho-phosphate from adsorption sites (ligand exchange andligand-promoted dissolution reactions), and throughchelation of cations (e.g. Al and Fe in acidic soils orCa in alkaline soils) that are commonly associatedwith orthophosphate in soil (Bar-Yosef 1991; Jones1998; Jones and Darrah 1994b). Organic anions alsomobilise P bound in humic-metal complexes (Gerke1993) and have been shown to increase both theavailability of organic P and its amenability todephosphorylation by phosphatases (Hayes et al.2000). However, the effectiveness of different organicanions in nutrient mobilization depends on variousfactors including; the form and amount of the

Distance from root surface (mm)

0 2 4 6 8 10 12 14

Soil pH

5.5

6.0

6.5= 0.09LSD0.05

z

900

1000

1100

1200

1300LSD 0.05 = 64.1

Acid phosphatase

270

275

280

285

290

LSD 0.05 = 7.0

NaOH-extractable organic P

(mg P

o kg-1

)

(mg C

kg-1 )

(g p-

NP g-

1h-

1 )

850

900

950

1000

1050

1100 LSD 0.05 = 56.1

200

250

300

350

400LSD 0.05 = 28.5

Microbial biomass carbon

(pH)

Water soluble organic carbon

0 2 4 6 8 10 12 14

130

140

150

LSD 0.05 = 4.3

(mg P

i kg-1

)

NaOH-extractable inorganic P(m

g C kg

-1 )

0 2 4 6 8 10 12 14

Fig. 2 Features of rhizosphere soil from perennial ryegrass(Lolium perenne L.; ..) and radiata pine (Pinus radiata D.Don; - -) when grown in a rhizobox system and comparedto an unplanted (control; __). Shown are changes inmicrobial biomass C, water-soluble C, pH, acid phosphataseactivity and NaOH-extractable inorganic and organic P contents

of the soil at various distances from the root surface. The soil(orthic brown soil; Dystrochrept; Hurunui, New Zealand) had atotal P content of 958 mg P kg1 soil. For each panel the errorbar (LSD P=0.05) shows least significant difference (data takenfrom Chen et al. 2002)

Plant Soil (2009) 321:305339 309

particular anion released, with citrate and oxalatebeing more effective relative to others (e.g. malate,malonate and tartrate followed by succinate, fumarate,acetate and lactate; Bar-Yosef 1991) and interactionsof the anion within the soil environment, including itseffective concentration in soil solution and relativeturnover rate (Jones 1998). The presence of micro-organisms is of further importance because of theircapacity to either rapidly metabolize different organicanions within the rhizosphere or through their ownability to release anions (Jones et al. 2003). Forexample, in a calcareous soil, Strm et al. (2001)showed greater stability and resistance to microbialdegradation of oxalate compared with citrate andmalate which resulted in greater mobilization of Pwithin localized regions of soil with a subsequentincrease in P uptake by maize (Zea mays L.) roots(Strm et al. 2002).

The effectiveness of organic anions in mobilizing Pfrom soil is highlighted by studies with white lupin(Lupinus albus L.) which exudes significant amountsof citrate (and to some extent malate) from cluster rootsthat are formed in response to P deficiency (Dinlelakeret al. 1989; Gardner et al. 1983; Keerthisinghe et al.1998; Neumann and Martinoia 2002; Vance et al.2003). Citrate is effective in mobilizing orthophosphatefrom pools of soil P that are otherwise not available toplants that either do not exude, or show limited releaseof organic anions, such as soybean (Glycine max (L.)Merr.) and wheat (Triticum aestivum L.) (Braum andHelmke 1995; Hocking et al. 1997). In addition, andanalogous to the role of root hairs, cluster roots have arelatively short life span and form on lateral roots asclosely packed tertiary roots with a dense covering ofroot hairs. This provides a zone for both concentratedrelease of organic anions and high surface area for theuptake of mobilized P (Dinkelaker et al. 1995,Neumann and Martinoia 2002). Interestingly, clusterroots form on plant species that are essentially non-mycorrhizal and therefore appear to provide animportant alternative strategy for plant acquisition ofsoil P (Shane and Lambers 2005). Indeed, theformation of cluster roots and high rates of organicanion release are reported for various native Australianspecies (e.g. the Proteaceae and Casuarinaceae fami-lies) which have evolved on low P soils (Roelofs et al.2001; Shane and Lambers 2005). Increased organicanion efflux from roots in response to P-deficiency alsooccurs in other species including chickpea (Cicer

arietinum L.) and pigeon pea (Cajanus cajan L.) andto a lesser extent in lucerne (alfalfa; Medicago sativaL.), canola (oil seed rape; Brassica napus L.) and rice(Oryza sativa L.) (Ae et al. 1991; Hedley et al. 1982;Hoffland et al. 1989; Lipton et al. 1987; Otani et al.1996; Pearse et al. 2006a; Veneklaas et al. 2003;Wouterlood et al. 2004). The increase in organic anionefflux by these species in response to P deficiencyhowever, is considerably less than for the Proteaceae orLupinus spp, and in many cases the agronomicsignificance of organic anion release remains to beverified in soil environments, as does the role ofvarious organic anions in mobilizing P from differentforms of soil P (Pearse et al. 2006b).

Activity of phosphatases is significantly greater inthe rhizosphere and is considered to be a generalresponse of plants to mobilize P from organic formsin response to P deficiency (Richardson et al. 2005).Phosphatases are required for the hydrolysis (miner-alization) of organic P, and in bulk soil microbial-mediated mineralization of organic P contributessignificantly to plant availability (Frossard et al.2000; Oehl et al. 2004). Depending on soil type andland management, organic forms of P commonlyconstitute around 50% of the total P in soil and is thepredominant form of P found in soil solutions (RonVaz et al. 1993; Shand et al. 1994). Dissolved organicP is derived largely from the turnover of soil micro-organisms and, relative to orthophosphate, has greatermobility in solution (Helal and Dressler 1989; Seelingand Zasoski 1993) and is therefore of criticalimportance to the dynamics and subsequent availabil-ity of P within the rhizosphere (Jakobsen et al. 2005a;Richardson et al. 2005).

Extracellular phosphatases released from rootshave been characterized for a wide range of plantspecies and been shown to be effective for the in vitrohydrolysis of various organic P substrates (Georgeet al. 2008; Hayes et al. 1999; Tadano et al. 1993;Tarafdar and Claassen 1988). Products of microbialturnover also contain high amounts of dissolvedorganic P (>80%) (primarily as nucleic acids andphospholipids), and are rapidly mineralized in soiland as such are of high availability to plants (Macklonet al. 1997). Direct hydrolysis of organic P andsubsequent utilization of released orthophosphate byroots has been demonstrated in soil using both wholeplant systems (e.g. McLaughlin et al. 1988) and inrhizobox studies. In the later case, depletion of

310 Plant Soil (2009) 321:305339

various pools of extractable organic P from therhizosphere is associated with higher activities ofphosphatases around plant roots (Chen et al. 2002;Tarafdar and Claassen 1988; Fig. 2). In the study byChen et al. (2002), greater net depletion of organic Pby radiata pine (compared to ryegrass) occurreddespite similar increase in phosphatase activity forboth species and the development of a lesser rootmat at the soil interface for pine roots, whichsuggests the possible involvement of other mecha-nisms. Indeed pine roots acidified the rhizosphere to agreater extent and had higher water soluble carbonand microbial biomass (Chen et al. 2002). Higherwater soluble C suggests either greater root exudationor higher turnover of microbial biomass which couldresult in increased P mobilization. Alternatively,greater radial depletion of P by pine roots maybeassociated with the presence of longer root hairs ormore likely by association with ectomycorrhizalfungi. This latter possibility is suggested by theauthors and is supported by observations from morerecent studies where mycorrhizal fungi have beenshown to be particular effective for the capture of Pby pine roots (Liu et al. 2005; Scott and Condron2004). Casarin et al. (2004) similarly showed theimportance of ectomycorrhizas for mobilization of poorlyavailable soil P around roots of maritime pine (Pinuspiaster Ait.) and that this benefit was from bothincreased soil exploration and, depending on the speciesof mycorrhizal fungi, due to the release of oxalate andprotons. However, the relative importance of suchmicrobial processes as compared to direct plant mech-anisms and other processes remains to be fullyestablished. In the study by Chen et al. (2002), numbersand activities of free-living and root-associated micro-organisms were also enhanced significantly within therhizosphere (e.g. as shown by increased microbialbiomass; Fig. 2) and these may also contributesubstantially to the mechanisms of P depletion andsolubilization.

Acquisition of nitrogen by plants

Nitrogen availability and uptake

Nitrogen occurs in soil in both organic and inorganicforms and in addition to marked seasonal changes ischaracterised by a heterogeneous distribution within thesoil. Nitrogen inputs through fixation reactions (by

either symbiotic microorganisms or potentially throughfree-living diazotrophs, as discussed below) and trans-formations of N between different pools have importantimplications for plant growth and for the loss of N fromsoil systems (Jackson et al. 2008). Microbial-mediatedmineralization of organic forms of N to ammonium(NH4

+) and its subsequent nitrification to nitrate(NO3

) is of major significance to N availability(Fig. 3) and has influence on root behaviour andrhizosphere dynamics. Although mineral forms of Nhave classically been considered to dominate plantuptake (see review by Miller and Cramer 2004), thereis evidence that soluble organic forms of N (e.g. lowmolecular weight compounds such as amino acids)may also play a significant role (Chapin et al. 1993),but few studies have quantified the relative importanceof each (Leadley et al. 1997; Schimel and Bennett2004). Of particular significance in the rhizosphere isthe effect that uptake of different N forms has on soilpH in the immediate vicinity of the root and subsequentinfluence of this on nutrient acquisition, especially inrelation to the availability of P and various micro-nutrients (e.g. Zn, Mn and Fe) (Marschner 1995).Changes in rhizosphere pH, caused by the influx ofprotons that occurs with uptake of NO3

, or the netrelease of protons for NH4

+ uptake, can also bringabout changes in the nature of substrates exuded fromroots or the quantities of exudates released, andconsequently may have major impact on the structureof microbial communities around the root (Bowen andRovira 1991; Meharg and Killham 1990; Smiley andCook 1983).

Both NO3 and NH4

+ reach the root surface via acombination of mass flow and diffusion (De Willigen1986). Nitrate is typically present in soil solution atmM concentrations and, relative to orthophosphate, ismore mobile (Tinker and Nye 2000) and thus, ispotentially able to move in soil by up to several mmper day (Gregory 2006). Ammonium is less mobilesince it readily adsorbs to the cation exchange sitesin soil and has lower rates for both mass flow anddiffusion. Nevertheless, diffusion and mass flow isthe major pathway for inorganic N uptake and,although it is difficult to differentiate diffusion fromroot interception, it is generally considered thatinterception of N in soil solution following rootextension accounts for a small percentage only of Ntaken up by plants (Barber 1995; Miller and Cramer2004).

Plant Soil (2009) 321:305339 311

Plant uptake of NH4+ and NO3

is a function oftheir concentrations in soil and soil solution, rootdistribution, soil water content and plant growth rate.The latter is most important under conditions ofliberal N supply, whereas mineral N concentrationand root distribution are more critical under Nlimiting conditions. Whilst some plant species showa preference for either NH4

+ or NO3 uptake, the

significance of this for N uptake at the field level isusually less than the abovementioned factors, espe-cially in agro-ecosystems (McNeill and Unkovich2007). Indeed, NH4

+ tends to dominate in manynatural ecosystems for reasons which may include asuppression of microbial-mediated nitrification(reviewed by Subbarao et al. 2006). In addition anddepending on soil type, non-exchangeable forms ofNH4

+ may contribute significantly to crop nutrition(Scherer and Ahrens 1996; Mengel et al. 1990),especially for lowland rice grown under floodedconditions (Keerthisinghe et al. 1985). Althoughmany crop plants differ in their sensitivity to toxiceffects of NH4

+, the crucial factor seems to be therelative concentration of the two ions, with theoptimal mix being dependent on factors such as plantspecies, age and soil pH (Britto and Kronzucker 2002;Badalucco and Nannipieri 2007). Furthermore, uptakeof NO3

across the plasma membrane is more costly

in terms of energy expenditure but nonetheless occurseffectively over almost the entire range of NO3

concentrations found in soil solutions (Forde andClarkson 1999). However, for crop production sys-tems the regulation of whole plant N uptake remainsrelatively poorly understood (Gastal and Lemaire2002; Jackson et al. 2008), although there is evidencethat the concentrations of both NO3

and NH4+ in soil

solution, as well as plant N status are involved (Aslamet al. 1996; Devienne-Barret et al. 2000).

The significance of soluble organic N for plantnutrition was first highlighted in solution culturestudies and has since been demonstrated for soils ina range of different ecosystems (Jones and Darrah1994a; Schimel and Bennett 2004). Amino acidstypically constitute up to half of the total soluble N insoil solution (concentrations ranging from 0.1 to50 mM) and thus comprise a significant part of thepotentially plant available N pool (Christou et al.2005; Jones et al. 2002). Amino acids in soil solutionoccur as a result of either direct exudation by roots orfrom the breakdown of proteins and peptides fromsoil organic matter and microbial biomass turnover asa result of microbial-derived proteases (Jaeger et al.1999; Owen and Jones 2001). In addition theinvolvement of plant-exuded proteases in digestionof proteins at the root surface has been reported along

diazotrophs

soil solution

DON organic N

NH3NH4+

NO3-

NH4+

NO3-

mineralization

immobilization

soil N

soil microorganisms

nitrification

mineral N

NO2- NO2-

NON2O

N2

plant N

N fertilizerplant residues, compost, animal excreta

soil organic matter

rootexudates

NH4+

NO3-

NO2-

uptake

uptake

organic N

assimilation

uptake

legumes

symbiotic N fixation

leaching

N2N2

NH3

denitrification

Fig. 3 The plant-soil Ncycle and pathways for Ntransformation mediated byphysiological processes(DON = dissolved organicnitrogen; redrawn fromMcNeill andUnkovich, 2007)

312 Plant Soil (2009) 321:305339

with the possibility of direct uptake of proteins byroots through endocytosis (Paungfoo-Lonhienne et al.2008). Uptake of organic N compounds by plantsmay also be facilitated by association with ectomy-corrhizal fungi (Chalot and Brun 1998; Nasholm et al.1998). However, the relative importance of suchmechanisms in many ecosystems remains debatablesince the diffusion rates of amino acids and proteinsare typically orders of magnitude lower than for NO3

(Kuzyakov et al. 2003), and thus mycorrhizas mayoffer distinct advantages, whereas microorganisms inthe rhizosphere are likely to compete more stronglyfor these compounds (Hodge et al. 2000a). Althoughdirect evidence that plant roots can out-competemicroorganisms for N is limited to a few studies(Hu et al. 2001; Jingguo and Bakken 1997), there issome evidence from 15N time-course studies whereplants accumulate 15N and thus benefit over thelonger term which may be associated with microbialturnover (Hodge et al. 2000b; Kaye and Hart 1997;Yevdokimov and Blagodatsky 1994).

Given the heterogeneous distribution of N in soil interms of chemical form, temporal dynamics andspatial distribution, plants adopt one or more of threemain strategies to optimize their acquisition of N.Broadly these are; (i) to explore greater volume of soiland/or soil solution by extending root length andbranching or increasing root surface area via changesin root diameter or root hair morphology, (ii) specificadaptive response mechanisms in order to exploitspatial and temporal niches such as N-rich patchesor due to the presence of particular forms of N (aminoacids, NH4

+ or NO3), and (iii) influences on plant

available N in the rhizosphere through plant-microbialinteractions.

Root growth and morphological responses to nitrogen

The size and architecture of the root system is animportant feature for ensuring adequate access to soilN, and root system size (relative to shoot growth) hasgenerally been shown to increase when N is limiting(Chapin 1980; Ericsson 1995). However, changes tobiomass alone are not necessarily indicative of thetotal absorptive area of a root system and morpho-logical changes can occur without change in biomass.Although the architecture of root systems is intrinsi-cally determined by genotype and the pattern of rootbranching, species specific attributes related to size

and architecture are also strongly determined byexternal physical, chemical and biological factors(Miller and Cramer 2004). Whilst primary rootgrowth is generally less sensitive to nutritional effectsthan is the growth of secondary or higher root orders(Forde and Lorenzo 2001), the diameter of first andsecond order laterals were significantly thicker incereals grown at high concentrations of NO3

(Drewet al. 1973). Thicker roots may be more costly toproduce but have greater capacity for the transport ofwater and nutrients and are less vulnerable to adverseedaphic conditions (Fitter 1987). Conversely, fineroots allow greater exploration of soil and plantsappear to accommodate trade-off between the two byexhibiting plasticity in root diameter (and morpholo-gy) according to the environmental conditions (Fordeand Lorenzo 2001). Root angle, an important com-ponent of root architecture in soil in relation to Pdeficiency (Rubio et al. 2003), appears to beunaffected by N deficiency.

Apart from the size and depth of root systems otherattributes may also influence the capacity for efficientcapture of N. Only a limited portion of the root mayactually be effective in the uptake of N (Robinson2001) and thus the spatial localization of roots isimportant when nutrient is distributed heterogeneous-ly (Ho et al. 2005). Electrophysiological and molec-ular evidence supports a role for root hairs in theuptake and transport of both NO3

and NH4+ (Gilroy

and Jones 2002) and proliferation of fine roots androot hairs in response to localized patches of N hasbeen demonstrated (Jackson et al., 2008).

Rooting depth, which varies greatly betweenspecies, influences the capture of N by plants,particularly of NO3

during periods of leaching, andis clearly an important characteristic for manyperennial agricultural species and tree crops (Gastaland Lemaire 2002). However, a dimorphic rootsystem, having both shallow and deep roots to enableacquisition of mineralized N in the topsoil as well asleached N at depth, is considered to be important (Hoet al. 2005). Indeed, vigorous wheat lines with fastervertical root growth and more extensive horizontalroot development have been shown to take upsignificantly more N (Liao et al. 2006).

Roots exhibit high plasticity as a physiologicalresponse to localized patches of organic and inorganicnutrients in soil, including proliferation in N-richzones (Hodge et al. 1999a; Robinson and van Vuuren

Plant Soil (2009) 321:305339 313

1998). This proliferation essentially involves theinitiation of new laterals, but may also includeincreases in the elongation rate of individual rootsand expansion of the rhizosphere through root hairs.Roots can also enhance their physiological ion-uptakecapacities in localized nutrient-rich zones. This rootforaging capability is considered to be an importantplant response to optimize resource allocation inregard to N capture and is of particular ecologicalimportance in situations where there is competitionwith neighbouring roots for limited resources (Hodgeet al. 1999b; Robinson et al. 1999). However,foraging is not a fixed property but varies withinspecies in response to different environmental con-ditions (Wijesinghe et al. 2001), indicating that theenvironmental context in which the root response isexpressed is as important as the response itself.However, most studies investigating root growth inresponse to patches of N have largely ignored theattributes of the patch itself despite the fact that thedynamics of nutrient transformation within the patchand microbial interactions are of major importance.There is need for research to follow both together,including consideration of the complexity of inter-actions with other root systems, soil microorganismsand fauna, and physical/chemical interactions in thesoil (Hodge 2004).

Association with microorganisms and plantgrowth promotion

Microbial associations with roots are complex in soiland can enhance the ability of plants to acquirenutrients from soil through a number of mechanisms.These include; i) an increase in the surface area ofroots by either a direct extension of existing rootsystems (e.g. mycorrhizal associations) or ii) byenhancement of root growth, branching and/or roothair development (e.g. through plant growth promot-ing rhizobacteria), (iii) a direct contribution to nutrientavailability though either N fixation (e.g. rhizobia anddiazotrophs) or by stimulation and/or contribution tometabolic processes that mobilize nutrients frompoorly available sources (e.g. organic anions) or, anindirect effect on nutrient availability by (iv) dis-placement of sorption equilibrium that results inincreased net transfer of nutrients into solution and/or as the mediators of transformation of nutrients

between different pools (e.g. nitrification inhibitorsand microbial-mediated processes that alter thedistribution of nutrients between inorganic and organ-ic forms) or v) through the turnover of microbialbiomass within rhizosphere (Gyaneshwar et al. 2002;Jakobsen et al. 2005a; Kucey et al. 1989; Richardson2007; Tinker 1980).

In this respect, mycorrhizal fungi, rhizobia andFrankia microsymbionts, and plant growth promoting(PGP) microorganisms are of particular importanceand as such have been studied most widely. PGPmicroorganisms represent a wide diversity of bacteriaand fungi that typically colonize the rhizosphere andare able to stimulate plant growth through either abiofertilizing (direct) effect or through mechanismsof biocontrol (indirect effect; Bashan and Holguin1997; and see Raaijmakers et al. 2009; Harman et al.2004; Fig. 4). Biofertilizing-PGPR (considered inmore detail below) specifically refers to rhizobacteriathat are able to promote growth by enhancing thesupply of nutrients to plants (Vessey 2003). Forrhizobia and Frankia this involves symbiotic relation-ships with host legume and actinorhizal plants,respectively (see review by Franche et al. 2009), andare therefore not considered here. Rhizobiaceae(rhizobia) can also develop non-specific associativeinteractions with roots and promote the growth ofnon-legumes, and as such are also commonly consid-ered as PGPR (Sessitsch et al. 2002). However, PGPis a complex phenomenon that often cannot beattributed to a single mechanism and, as outlinedbelow, PGPR may typically display a combination ofmechanisms (Ahmad et al. 2008; Kuklinsky-Sobral etal. 2004). In addition, the effects of individual PGPRmay not occur alone but through synergistic inter-actions between different microorganisms.

Association with mycorrhizal fungi

Mycorrhizal symbioses are found in almost allecosystems and can enhance plant growth througha number of processes which include improvementof plant establishment, increased nutrient uptake(particularly P and essential micronutrients such asZn and Cu, but also N and, depending on soil pH,may enhance the uptake of K, Ca and Mg; Clarkand Zeto 2000), protection against biotic and abioticstresses and improved soil structure (Buscot 2005;Smith and Read 2008). Mycorrhizal fungi typically

314 Plant Soil (2009) 321:305339

colonize the root cortex biotrophically and developexternal hyphae (or extra-radical mycelia) whichconnect the root with the surrounding soil. All butfew vascular plant species are able to associate withmycorrhizal fungi. The universality of this symbiosisimplies a great diversity in the taxonomic features ofthe fungi and the plants involved. At least five types

of mycorrhizas are recognized, the structural andfunctional features of which are reviewed in detailelsewhere (Smith and Read 2008; Brundrett, 2002)and are thus only considered briefly here.

Higher plants commonly form associations withectomycorrhizas, mainly forest trees in the Fagaceae,Betulaceae, Pinaceae, Eucalyptus, and some woody

NO

ACC deamination

Pathogens

AM fungi

Biocontrol-PGPR

AntagonismCompetition

Biofertilizing-PGPR

Denitrification

Free N2fixationNO

N2ON2

NH4+

N03-NO2 -

Nitrification

Production of phytohormones,

siderophores, vitaminsorganic P

inorganic P

Mineralizationphosphate

Solubilization

Ethylene

IAA

ACC ACC NH3+ -KB

Biofertilizing-PGPR

Biofertilizing-PGPR

P nutrition

N nutritionBNI

Elicitingplant defense responses

positive effects of plant regulators on root development or on hormonal pathways

negative effects of plant regulators on root development or on hormonal pathway

positive effects of microorganisms on plant nutrition and/or root development

plant exudates

microbial enzymatic processes

negative effects of PGPR or plants on detrimental rhizospheric microorganisms or microbial processes

negative effects of microorganisms on plant nutrition and/or root development

Interactions of plant growthpromoting rhizobacteria

(PGPR)

Fig. 4 Plant growth promotion mechanisms (positive andnegative effects) associated with soil and rhizosphere (PGPR)microorganisms. Biofertilizing-PGPR and arbuscular mycorrhi-zal (AM) fungi stimulate plant nutrition by directly increasingthe supply of nutrients to plants (e.g. through N fixation, Psolubilization and/or mineralization, vitamin and siderophoreproduction) or by increasing the plants access to nutrients due

to enhancement of root volume. Promotion of root growth islinked to the ability of PGPR to produce phytohormones (e.g.IAA, ethylene, NO) or by direct influences on plant hormonelevels (e.g. deamination of ACC precursor to plant ethylene).Biocontrol-PGPR improve plant heath by inhibiting the growthof plant pathogens or by eliciting plant defense responses

Plant Soil (2009) 321:305339 315

legumes. The fungi involved are usually Basidiomy-cetes and Ascomycetes which colonize cortical roottissues but without intracellular penetration (Smithand Read 2008). Three other types of mycorrhizas canbe grouped as endomycorrhizas, in which the funguscolonizes the root cortex intercellularly. One of theseis restricted to species within the Ericaceae (ericoidmycorrhizas), the second to the Orchidaceae (orchidmycorrhizas) and the third, which is by far the mostwidespread (and therefore considered in most detailhere), are the arbuscular mycorrhizas (AM). A fifthgroup, the ectendomycorrhizas, are associated withplant species in families other than Ericaceae,including the Ericaless and the Monotropaceae (arbu-toid and monotropoid mycorrhizas). The majority ofplant families form arbuscular associations, with theAM fungi being an obligate symbiont that is unable tocomplete its life cycle without colonization of a hostplant. The AM fungi were formerly included in theorder Glomales, Zygomycota, but are now consideredas new phylum, the Glomeromycota (Redecker 2002,Schbler et al. 2001).

The establishment of mycorrhizal fungi in rootschanges key aspects of plant physiology, includingmineral nutrient composition in tissues, plant hor-monal balance and patterns of C allocation. The fungimay also alter the chemical composition of rootexudates, whilst the development of mycelium in soilcan act as a C source for microbial communities andintroduce physical modifications to the soil environ-ment (Gryndler 2000). Such changes in the rhizo-sphere can affect microbial populations bothquantitatively and qualitatively such that the rhizo-sphere of mycorrhizal plants (known as the mycor-rhizosphere) generally has features that differsubstantially from those of non-mycorrhizal plants(Barea et al. 2002; Johansson et al. 2004; Offre et al.2007). As discussed in more detail below, a widerange of bacteria (including actinomycetes and vari-ous PGPR) associate with mycorrhizas within themycorrhizosphere (Rillig et al. 2006; Toljander et al.2007).

Contribution of mycorrhizas to the phosphorusnutrition of plants

It is well established that mycorrhizal fungi contributesignificantly to the P nutrition of plants, particularlyunder low P conditions (Barea et al. 2008). This is

most evident for the ectomycorrhizal fungi which arelargely associated with non-agricultural plants andappear to show greater functional diversity than theAM fungi (Brundrett 2002). Whilst it is generallyaccepted that mycorrhizal fungi have similar access tosources of P in soil solution that are also directlyavailable to plants (reviewed by Bolan 1991) thereis some evidence to suggest that both AM andectomycorrhizal fungi have enhanced ability to usealternative sources of P (Bolan et al. 1984; Casarin etal. 2004). For example, for AM fungi, Tawaraya et al.(2006) showed that exudates from fungal hyphaesolubilized more P than root exudates alone, suggest-ing that the mycorrhiza contribute to increased Puptake through solubilization. The extra-radical my-celium of AM fungi have also been shown to excretephosphatases which could potentially enhance themineralization and utilization of organic P (Koide andKabir 2000; Joner and Johansen 2000). However, it isgenerally considered that this is unlikely to be ofmajor significance to the overall contribution of AMfungi to plant P nutrition (Joner et al. 2000,Richardson et al. 2007). Alternatively, indirect effectsof mycorrhizal fungi on P mobility may occurthrough changes in soil microbial communities withinthe mycorrhizosphere (Barea et al. 2005b).

The increased efficiency of P acquisition bymycorrhizal plants is based mainly on the existenceof the extra-radical mycelia which develop into soiland allow P to be accessed by the mycorrhiza fromsoil solution at distances up to several cm away fromthe root and then subsequently transferred to the plant(Jakobsen et al. 1992). High mycorrhizal hyphaedensity also provides considerably greater surfacearea for the absorption of orthophosphate by plantsand, due to the smaller size of hyphae in relation toroots and root hairs and their greater length relative toroot hairs, hyphae are also most effective in exploitingsoil pores and nutrient patches that may not bedirectly accessible to roots (Jakobsen et al. 2005a).Thus, higher uptake of P by mycorrhizal plants (bothectomycorrhizas and AM) can generally be explainedin term of increased hyphal exploitation of the soil asmodelled by Schneph et al. (2008) and the compet-itive ability of the hyphae to absorb localized sourcesof orthophosphate and organic nutrient patches(Tibbett and Sanders 2002; Cavagnaro et al. 2005).In this context, numerous studies have shown positivecorrelations between fungal variables, such as hyphal

316 Plant Soil (2009) 321:305339

length or hyphal density, with growth responsevariables of colonized plants such as shoot biomass,P uptake and total P content (Avio et al. 2006;Jakobsen et al. 2001). However, this cannot be takenas a general conclusion since high hyphal develop-ment does not always correlate with plant growthresponses (Smith et al. 2004) and in some situations,particularly under fertilized field conditions, thepresence of AM mycorrhizal fungi appears to providelittle or no benefit in terms of plant P nutrition (Ryanand Angus 2003, Ryan et al. 2005).

Apart from the physical extension of root systems,mycorrhizal fungi may also acquire orthophosphatefrom soil solution at lower concentrations than roots,but whether this contributes significant advantage tothe P nutrition of plants remains uncertain. Somestudies report higher affinity for orthophosphateuptake by mycorrhizal plants (i.e. lower Km valuesthan for non-mycorrhizal roots; Cardoso et al. 2006).Genes encoding for the high-affinity phosphatetransport in AM fungi have been identified andshown to be preferentially expressed in the extra-radical mycelium (Benedetto et al. 2005; Maldonado-Mendoza et al. 2001). Mycorrhizal plants thereforehave two pathways for P uptake, the directpathway via the plant-soil interface through roothairs, and the mycorrhizal pathway via the fungalmycelium (Smith et al. 2003). Interestingly, severalstudies have shown that expression of plant epidermalP transporters is reduced in roots that are colonized byAM fungi, and that under these circumstances Puptake proceeds predominantly via fungal transport-ers with subsequent transfer of P to plants at thearbuscular-root interface (Burleigh et al. 2002; Chiouet al. 2001; Liu et al. 1998; Rausch et al. 2001). Insome cases AM colonization results in a completeinactivation of the direct P uptake pathway via roothairs with essentially all of the P in plant tissues beingprovided through the mycorrhizal route (Smith et al.2004).

Contribution of mycorrhizas to nitrogen nutrition

Several studies have also shown increased N uptakefrom soil by roots associated with mycorrhizal fungi(Barea et al. 2005a). For example, Ames et al. (1983)first showed that mycorrhizal hyphae were able toabsorb, transport and utilize NH4

+ and, using 15N-based techniques, Barea et al. (1987) demonstrated

that mycorrhizal plants under field conditions hadincreased N uptake. Further studies using 15N andcompartmented rhizobox systems verified that my-corrhizal hyphae in root-free compartments were ableto access 15N (Tobar et al. 1994b). To investigatewhether the mycorrhizal contribution to N acquisitionwas from pools of N that were unavailable to nonmycorrhizal plants, the apparent pool size of plantavailable N has been determined by isotopic dilution(i.e. the AN value of the soil; Zapata 1990). Using thisapproach, higher AN values for plants inoculated withAM fungi compared to non-inoculated controls wereobtained, suggesting that the AM mycelium were ableto access N from forms that were otherwise lessavailable than that for non-mycorrhizal plants (Bareaet al. 1991). Indeed the presence of a functionaltransporter for high-affinity uptake of NH4

+ hasrecently been identified in the extraradical myceliumof Glomus sp. (Lpez-Pedrosa et al. 2006).

Association with free-living microorganisms

Phosphate-mobilizing microorganisms

Free-living bacteria and fungi that are able to mobilizeorthophosphate from different forms of organic andinorganic P have commonly been isolated from soil andin particular from the rhizosphere of plants (Kucey et al.1989; Barea et al. 2005b). Phosphate-solubilizingmicroorganisms (PSM) are characterized by theircapacity to solubilize precipitated forms of P whencultured in laboratory media and include a wide rangeof both symbiotic and non-symbiotic organisms, suchas Pseudomonas, Bacillus and Rhizobium spp., actino-mycetes and various fungi such as Aspergillus andPenicillium spp. (see reviews by Gyaneshwar et al.2002; Kucey et al. 1989; Rodriguez and Frago 1999;Subba-Rao 1982; Whitelaw 2000). Selection of PSM isroutinely based on the solubilization of sparinglysoluble Ca phosphates (typically, tri-calcium phosphate[Ca3(PO4)2] and rock phosphates containing hydroxy-and fluor-apatites [Ca5(PO4)3OH and Ca10(PO4)6F2])and Fe and Al phosphates such as strengite(FePO4.2H2O) and variscite (AlPO4.2H2O). Theamount of P solubilized is highly dependent on thesource (solubility) of the P and, for different micro-organisms, is influenced to a large extent by the cultureconditions. For example, fungi are commonly reportedto be more effective at solubilization of Fe and Al

Plant Soil (2009) 321:305339 317

phosphates, whereas the ability of different organismsto solubilize Ca-phosphates is influenced by the sourceof carbon and nitrogen in the media, by the bufferingcapacity of the media and the stage at which culturesare sampled (Kucey 1983; Illmer and Schinner 1995;Nahas 2007; Whitelaw et al. 1999). From variousstudies it is evident that change in pH of the mediais particularly important for solubilization of Ca-phosphates, whereby cultures supplied with NH4

+ aremore effective than those with NO3

- due to associatedproton release and acidification of the media. Acidifi-cation is also commonly associated with the release oforganic anions which have been widely reported forvarious microorganisms (i.e. with citrate, oxalate,lactate, and gluconate being most common). Organicanions themselves may further increase the mobiliza-tion of particular forms of poorly soluble P (e.g. Al-Pand Fe-P) through chelation reactions (Whitelaw2000).

Under controlled growth conditions various studieshave demonstrated enhanced growth and P nutritionof plants inoculated with PSM which is oftenattributed to the P-solubilizing activity of the micro-organisms involved (see reviews by Gyaneshwar et al.2002; Kucey et al. 1989; Rodriguez and Fraga 1999;Whitelaw 2000). However, clear effect of PSM inmore complex soil environments and in field con-ditions, have proved more difficult to demonstrate andinconsistent response of plants and performance ofdifferent microorganisms have been observed. Asdiscussed by Richardson (2001) this may be due toa range of factors that include, insufficient knowledgefor introducing and understanding the dynamics ofmicroorganisms and their interaction with complexmicrobial communities in soil, the apparent lack ofany specific association between partners, and poorunderstanding of the actual mechanisms involved,both for the microorganisms and their interaction andefficacy within different soil environments. Forexample, whilst Penicillium radicum effectively sol-ubilized P in laboratory media and was able topromote the growth of wheat, evidence for improvedP uptake in response to inoculation was only evidentin glasshouse trials particularly where fertilizer P wasapplied (Whitelaw et al. 1997). However, growthpromotion may not necessarily be directly associatedwith P solubilization and production of phytohor-mones is likely to be involved (Wakelin et al. 2006).Similarly, promotion of root growth and enhanced P

nutrition of plants inoculated with P. bilaii has beenshown to be primarily associated with increased rootgrowth, including greater specific root length andproduction of longer root hairs (Gulden and Vessey2000). Such studies highlight the difficulties indetermining the actual mechanism associated withgrowth promotion, as stimulation of root growth alsocontributes to greater potential for P acquisition. It isimportant therefore that experiments directed atdemonstrating the benefits of PSM be conductedacross a range of P supplies, whereby benefits ofinoculation should be negated at higher levels ofapplied P, and that specific measures of P acquisition(e.g. by isotopic dilution) be made to confirm Pmobilization from pools of soil that are otherwisepoorly available to roots.

The mineralization of organic P in soil is largelymediated by microbial processes and as such micro-organisms play a significant role in maintaining plantavailable P. Microorganisms are able to hydrolyze awide range of organic P substrates when grown inculture and, when added to soil, different forms oforganic P have been shown to be rapidly mineralized(Adams and Pate 1992; Macklon et al. 1997; and seereview by Richardson et al. 2005). Indeed benefits ofmicrobial inoculation for the utilization of organic Pby plants under controlled growth conditions havebeen demonstrated (Richardson et al. 2001a). Inaddition, the microbial biomass is important formaintaining both inorganic and organic P in soilsolution (Seeling and Zososki 1993) and turnover ofthe biomass represents an important potential supplyof P to plants (Oberson and Joner 2005). Thiscontribution is likely to be of greater significance inthe rhizosphere where there is increased amount ofreadily metabolizable carbon and higher density ofmicroorganisms (Brimecombe et al. 2007; Jakobsenet al. 2005b). However, the relative importance ofmicrobial mineralization relative to the short-termimmobilization of P by microorganisms in therhizosphere and its impact on the availability oforthophosphate to plants requires more detailedinvestigation.

Whilst it is evident that microbial-mediated solu-bilization and mineralization of inorganic and organicP are important processes whereby microorganismsare able to acquire P from soil, it has been argued thatthey are unlikely to mobilize sufficient P above theirown requirements to meet plant demand (Tinker

318 Plant Soil (2009) 321:305339

1980). Indeed, few studies have unequivocally dem-onstrated a direct release of P by microorganisms insoil and benefits to plant nutrition are therefore ofteninferred. Nevertheless, the cycling of P within themicrobial biomass and its subsequent release isparamount to the P cycle in soil and represents animportant pathway for movement of P from varioussoil pools into plant-available forms and may alsoserve to protect orthophosphate from becomingunavailable in soil due to various physicochemicalreactions (Magid et al. 1996; Oberson et al. 2001).The significance of this in the rhizosphere warrantsfurther research.

Microbial interactions and nitrogen availability

Root exudation also has important implications forN availability. Although the chemical compositionof exudates varies widely for different species androot types and is primarily comprised of Ccompounds, exudates can also contain significantquantities of N, which is either available tomicroorganisms in the rhizosphere or can berecaptured by plants (Bertin et al. 2003; Uren2007). In addition, root exudates are the majorenergy supply for the soil food web and play asignificant role in the turnover of soil organic matterand associated nutrients. However, as suggested byJones et al. (2004), although root exudates have beenhypothesized to be involved in the enhanced mobi-lization and acquisition for many nutrients in soil,there is little mechanistic evidence from soil-basedsystems to verify this, which is further highlightedby a recent analysis of the literature concerningrhizodeposition by maize (Amos and Walters 2006).Despite this, some studies have demonstrated en-hanced N cycling in the vicinity of plant roots(Jackson et al. 2008). For example, following theapplication of 15N labelled fertilizer, the excess of15N in the microbial biomass increased significantlyin both planted and control soils at up to 8 weeksafter plant emergence, but then declined in controlsoils only (Qian et al. 1997). Retention of 15N in themicrobial biomass in planted soil was attributed tothe release of root-derived C from maize as estimat-ed using 13C abundance methodology. This releaseof exudate was suggested to promote microbialimmobilization of the N and is consistent withgreater microbial biomass in the rhizosphere.

Increased rates of N mineralization have similarlybeen demonstrated in the rhizosphere of slender wildoats (Avena barbata Pott ex Link), where N mineral-ization was 10 times higher than in bulk soil, but thiswas highly dependent on location along the root(Herman et al. 2006). In addition, a rhizospherepriming effect has been suggested to be involved inthe decomposition of native soil organic matter aroundroots (Cheng et al. 2003; Kuzyakov 2002). However,stimulation of mineralization may be dependent onplant species and on the C:N ratio of substrates ashighlighted in a study of peas (Pisum sativum L.)inoculated with Pseudomonas fluorescens, whereincreased uptake of N from 15N enriched organicresidues occurred, whereas decreased uptake wasobserved for wheat (Brimecombe et al. 1999). Addi-tional work demonstrated that microbial-microfaunalinteractions in the rhizosphere were also involved inthis differential response, with lower numbers ofnematodes and protozoa being present in the rhizo-sphere of uninoculated peas which appeared to exert anematicidal effect (Brimecombe et al. 2000). On thecontrary, higher numbers of nematodes and protozoa(e.g. up to 6-fold) have generally been reported in therhizosphere where they specifically feed on bacteria,fungi and yeasts (Zwart et al. 1994). These interactionsenhance N flows in the rhizosphere both directly, viathe excretion of consumed nutrients and mineralizationof nutrients on death (Griffiths 1989), and indirectlyvia changes to the composition and activity of themicrobial community (Griffiths et al. 1999). Forexample, bacteriophagous nematodes mineralized upto six times more N than an equivalent biomass ofprotozoa grazing on bacteria (Griffiths 1990). Netmineralization is due to differences in C:N ratiobetween the protozoan (or nematode) predator andthe bacterial prey and their relatively low assimilationefficiency, whereby ~60% of ingested nutrients aretypically excreted and thus potentially available forplant or microbial uptake (Bonkowski 2004). More-over, plants grown under controlled conditions havebeen observed to develop more highly branched rootsystems in the presence of protozoa which may partlybe explained as a response to NO3

formed frommineralization of NH4

+ excreted by protozoa (Forde2002). However, other work has suggested that rootresponses are due to a direct phytohormone effect bythe presence of either protozoa or a consequence ofauxin producing bacteria stimulated by the presence of

Plant Soil (2009) 321:305339 319

the protozoa (Bonkowski 2004; Bonkowski and Brandt2002). The consequences of soil organisms promotinga mutually beneficial relationship between plant rootsand bacteria in the rhizosphere on root architecture,nutrient uptake and plant productivity is therefore ofcurrent research interest (Mantelin and Touraine 2004).For example, in an experiment using 15N/13Clabelled organic nutrient sources combined withmanipulation of the composition of microfaunal pop-ulations, Bonkowski et al. (2000) observed effects onryegrass growth and concluded that microfaunalgrazing increased the temporal coupling of nutrientrelease and plant uptake, whereas root foraging inorganic nutrient-rich zones enhanced the spatial cou-pling of mineralization and plant uptake. More recentwork has shown interaction at higher trophic levelswhere collembola that feed on bacteria, fungi, nemat-odes and protozoa further enhanced N mineralizationwithout alteration to the microbial biomass C (Kanedaand Kaneko 2008).

At a larger scale there is increasing evidence fromusing in situ 15N labelling of plant root systems of theimportance of N rhizodeposition in sustaining the Ncycle of agro-ecosystems (Hogh-Jensen 2006; Mayeret al. 2004; McNeill and Fillery 2008). A recent review(Wichern et al. 2008) highlights the wide variability inresults and suggests there is need for more inves-tigations on key environmental factors influencing theamounts of N released under field conditions fromdifferent species. Apart from a direct influence of therhizosphere on N deposition, there is also need to morefully understand the role that plant roots have ininteracting with soil microorganisms and influencingother parts of the soil N cycle. For example, thepresence of plant-derived biological nitrification inhib-itors (BNI) in the root zone can influence theconversion of NH4

+ to NO3- (nitrification) and subse-

quently to the potential for gaseous losses of N throughdenitrification (Fig. 3). Indeed nitrification inhibitorshave been recognized for some time in native plantecosystems, but more recently have been shown to beeffective as root exudates from brachiaria (Brachiariahumidicola (Rendle) Schweick) a tropical grass species(Subbarao et al. 2006). In these systems there is greaterretention of NH4

+ in soil which has importantimplication for improving N-use efficiency by reducingpotential N losses through NO3

leaching and/or itsconversion (denitrification) to N2O gas (via NO),which contributes significantly to greenhouse-gas

emissions (Fillery 2007; Subbarao et al. 2006).However, presently it appears that BNI is poorlyexpressed in many agricultural crop species, althoughactivities has been reported for sorghum (Sorghumbicolour (L.) Moench.), pearl millet (Pennisetumglaucum (L.) R.Br.) and peanut (Arachis hypogaeaL.) (Subbarao et al. 2007a) and more recently BNI hasbeen shown to be effective in wild rye (Leymusracemosus (Lam.) Tzvelev), a wild relative of wheat(Subbarao et al. 2007b).

Nitrogen fixation by diazotrophs

Diazotrophs (i.e. N2-fixing bacteria) are classified asbeing either symbiotic (rhizobia and Frankia species)or as free-living (associative) and/or root endophyticmicroorganisms (Cocking 2003). Rhizobia developsymbiotic relationships with host legumes andthrough atmospheric N2 fixation within nodules canprovide up to 90% of the N requirements of the plant(see Franche et al. 2009; Hflich et al. 1994). Free-living N2-fixers also have the potential for providingN to host plants but so far, the direct contribution ofN-fixation by diazotrophs to the N nutrition of plantsand subsequent growth promotion remains in ques-tion. Free-living diazotrophs have been identified inseveral genera of common rhizosphere-inhabitingmicroorganisms such as Acetobacter, Azoarcus, Azo-spirillum, Azotobacter, Burkholeria, Enterobacter,Herbaspirillum, Gluconobacter and Pseudomonas(Baldani et al. 1997; Mirza et al. 2006; Vessey2003), with some being recognized as endophytes.Endophytic diazotrophs may have advantage overroot-surface associated organisms, as they can colo-nize the interior of plant roots and establish them-selves within niches that are more conducive toeffective N2 fixation and subsequent transfer of thefixed N to host plants (Baldani et al. 1997; Reinhold-Hurek and Hurek 1998).

Mutants deficient in nitrogenase activity (i.e. Nif-

mutants) have been constructed in various PGPR,including Azospirillum brasilense, Azoarcus sp. andPseudomonas putida, and importantly, have beenshown in several cases to retain their ability to promoteplant growth of certain crops (Hurek et al. 1994;Lifshitz et al. 1987). This questions the relativecontribution of N2 fixation to the growth promotioneffect. On the contrary, Hurek et al. (2002) showed thatan endophytic strain of Azoarcus sp. was able to fix

320 Plant Soil (2009) 321:305339

and transfer N when associated with kallar grass(Leptochloa fusca (L.) Kunth), as Nif- mutants gavelower plant growth stimulation than the wild-typestrain. Similar results were obtained in the case of theassociative symbiosis between sugar-cane (Saccharumofficinarum L.) and the endophytic diazotroph Aceto-bacter diazotrophicus (renamed Gluconobacterdiazotrophicus) (Sevilla et al. 2001). Furthermore, N-balance, 15N natural abundance and 15N dilutionstudies, performed in either pot experiments or in thefield, have provided clear evidence of the ability ofendophytic N2-fixing bacteria to supply significantinputs of nitrogen to some grasses and cereals (Boddeyet al. 2001; Oliveira et al. 2002).

In contrast to symbiotic N2 fixation, where there isdirect transfer of N across the symbiotic interface, it isevident that root surface associated diazotrophs seemnot able to readily release fixed N to the host plant andthat this occurs only through microbial turnover(Lethbridge and Davidson 1983; Rao et al. 1998).This may account for inconsistent response of plantsinoculated with diazotrophs and indicates that there isneed for better understanding of the potential for free-living and endophytic diazotrophs to supply N to hostplants. In addition to their potential for supplyingplants with N, free-living diazotrophs may alsopromote plant growth and nutrition through variousother mechanisms.

Other mechanisms of PGPR to enhance plantnutrition

Microbial production of phytohormones

Many PGPR produce phytohormones that areconsidered to enhance root growth and greatersurface area (e.g. bigger roots, more lateral rootsand root hairs) leading to an increase in exploredsoil volume and thus plant nutrition (Fig. 4). Suchmicroorganisms, commonly termed phytostimula-tors, include a wide range of soil bacteria and fungi.The most common phytohormones produced byPGPR are auxins, cytokinins, giberellins and to alesser extent ethylene, with auxins being the mostwell characterized (Khalid et al. 2004; Patten andGlick 1996; and see review by Arshad andFrankenberger 1998). Indeed in Azospirillum, auxin(IAA) production, rather than N2 fixation, is gener-ally considered to be the major factor responsible for

the PGPR response through stimulation of rootgrowth (Dobbelaere et al. 1999).

Indole-3-acetic acid (IAA) controls a wide varietyof processes in plant development and plant growthand plays a key role in shaping plant root architecturesuch as regulation of lateral root initiation, rootvascular tissue differentiation, polar root hair posi-tioning, root meristem maintenance and root gravitro-phism (Aloni et al. 2006; Fukaki et al. 2007).Production of IAA is widespread among rhizobacteria(Khalid et al 2004; Patten and Glick 1996; Spaepen etal. 2007), with increasing numbers of endophyticIAA-producing PGPR being reported (Tan and Zou2001). For example, Kuklinsky-Sobral et al. (2004)screened a collection of root-associated bacteria fromsoybean for their ability to produce IAA and showedthat it was present in 28% of isolates. More recently,the distribution of IAA biosynthetic pathways amongannotated bacterial genomes suggests that 15% (from369 analysed) contain genes necessary for synthesisof IAA (Spaepen et al. 2007). IAA production hasalso been observed in rhizobia and in phytopathogen-ic bacteria, although the amount of auxins producedby different rhizobacteria seems to differ according totheir mode of interaction with plants (Kawaguchi andSyno 1996). Several IAA biosynthetic pathways,classified according to their intermediates, exist inbacteria and for most, tryptophan has been identifiedas the precursor of IAA (Patten and Glick 1996;Spaepen et al. 2007). However, only few specific genesand proteins involved in IAA biosynthesis have beencharacterized to date, and only in a small number ofPGPR (e.g. Azospirillum brasilense, Enterobactercloacae, Pantoea agglomerans and Pseudomonasputida; Koga et al. 1991; Patten and Glick 2002;Zimmer et al. 1998). In phytopathogenic bacteria, IAAseems to be mainly produced from tryptophan via theintermediate indole-3-acetamide (IAM pathway),whilst in beneficial phytostimulatory bacteria, IAAappears to be synthesized predominantly via indole-3-pyruvic acid (IPyA pathway) (Manulis et al. 1998;Patten and Glick 1996; 2002; Zimmer et al. 1998).

In A. brasilense, inactivation of a key enzyme inthe IPyA pathway (the ipdC gene, encoding anindole-3-pyruvate decarboxylase) resulted in up to90% reduction of IAA production (Dobbelaere et al.1999), but mutants were not completely abolished inIAA biosynthesis, suggesting some redundancy inpathways. Irrespective of this, various ipdC mutants

Plant Soil (2009) 321:305339 321

displayed altered phenotypes compared to the wild-type strains in their ability to alter wheat rootmorphology (i.e. either no increase in root hair andlateral root formation and no decrease in root length;Malhotra and Srivastava 2008; Dobbelaere et al1999). The impact of exogenous auxin on plantdevelopment ranges from positive to negative effects,and occurs as a function of the amount of IAA produced,the cell number of auxin-producing rhizobacteria and onthe sensitivity of the host plant to changes in IAAconcentration (Dobbelaere et al 1999; Spaepen et al2008; Xie et al. 1996). For example, Remans et al.(2008) highlighted cultivar specificity in the responseof plants to auxin-producing bacterial strains. In manyPGPR, genes involved in IAA production are fine-regulated by stress factors that commonly occur in soiland potentially within the rhizosphere (e.g. acidic pHand osmotic stress), and in some cases have beenshown to be activated by plant extracts (e.g. aminoacids such as tryptophan, tyrosine and phenylalanine,auxins and flavonoids; Ona et al. 2005; Prinsen et al.1991; Vande Broek et al. 1999; Zimmer et al. 1998).

Cytokinins stimulate plant cell division, control rootmeristem differentiation, inhibit primary root elongationand lateral root formation but can promote root hairdevelopment (Riefler et al. 2006; Silverman et al.1998). Cytokinin production has been reported invarious PGPR including, Arthrobacter spp., Azospir-illum spp., Pseudomonas fluorescens, and Paenibacil-lus polymyxa (Cacciari et al. 1989; de Salamone et al.2001; Perrig et al. 2007; Timmusk et al. 1999).However, genes involved in the biosynthesis ofbacterial cytokinins have not yet been characterizedin PGPR and therefore their involvement in plantgrowth promotion largely remains speculative.

Gibberellins enhance the development of planttissues particularly stem tissue and promote rootelongation and lateral root extension (Barlow et al.1991; Yaxley et al. 2001). Production of gibberellinshave been documented in several PGPR such asAzospirillum spp., Azotobacter spp., Bacillus pumilus,B. licheniformis, Herbaspirillum seropedicae, Gluco-nobacter diazotrophicus and rhizobia (Bottini et al.2004; Gutirrez-Maero et al. 2001). Some Azospir-illum strains are also able to hydrolyze, both in vitro(Piccoli et al. 1997) and in vivo (Cassn et al. 2001),glucosyl-conjugates of gibberellic acid, which corre-spond to reserve or transport forms of gibberellic acidproduced by plants (Schneider and Schliemann 1994).

This activity leads to an increase in the release ofactive forms of the phytohormone into the rhizosphere.However, the bacterial genetic determinants involvedin this mechanism remain to be identified as does theprecise role of gibberellins in plant growth promotionby PGPR.

Ethylene is a key phytohormone that can inhibitroot elongation, nodulation and auxin transport, andpromotes seed germination, senescence and abscis-sion of various organs and fruit ripening (Bleeckerand Kende 2000; Glick et al. 2007b). Ethylene isrequired for the induction of systemic resistance inplants during associative and symbiotic plant-bacteriainteractions and, at higher concentrations, is involvedin plant defence pathways induced in response topathogen infection (Broekaert et al. 2006; Glick et al.2007a). Certain PGPR such as A. brasilense havebeen shown to produce small amounts of ethylenefrom methionine as a precursor (Perrig et al. 2007;Thuler et al. 2003), and this ability seems to promoteroot hair development in tomato plants (Ribaudo et al.2006). However, a better knowledge (i.e. character-ization of bacterial biosynthesis pathway and geneticdeterminants involved) has to be gained in order todetermine the role of the production of this plantgrowth regulator in the growth promoting effect ofPGPR.

Some plant-associated bacteria such as A. brasi-lense (strain Sp245) are able to produce nitric oxide(NO) due to the activity of nitrite reductases (Creus etal. 2005; Pothier et al. 2007). The formation of NO isan intermediate in the denitrification pathway, duringwhich nitrate (or nitrite) is converted to nitrogenoxides (N2O) and to N2 (Zumft 1997; Fig. 3). Thispathway is utilized by soil bacteria to gain energyunder oxygen-limited conditions that may occur in therhizosphere (Hjberg et al. 1999). Although denitri-fication by rhizobacteria diminishes the amount ofNO3

available for plant nutrition, it may havepositive effects on root development by means ofNO production, which is a key signal molecule thatcontrols root growth and nodulation, stimulates seedgermination and is involved in plant defenceresponses against pathogens (Lamattina et al. 2003;Pii et al. 2007). Furthermore, NO can interact withother plant hormone signalling networks includingthat for IAA (Lamattina et al 2003; Fig. 4). Bacterialdenitrification and production of NO by A. brasilensehas been demonstrated on wheat roots (Creus et al.

322 Plant Soil (2009) 321:305339

2005; Neuer et al. 1985; Pothier et al. 2007) and NOproduced during tomato root colonization stimulatedthe formation of lateral roots (Creus et al. 2005).Nitrous oxide production therefore is potentiallyanother plant-beneficial trait displayed by Azospiril-lum PGPR.

Microbial enzymatic activities influencing planthormone levels

Certain PGPR are able to stimulate plant growth bydirectly lowering plant ethylene levels through theaction of 1-aminocyclopropane-1-carboxylic acid(ACC) deaminase (i.e. deamination of the plant ethyleneprecursor; Fig. 4). ACC deaminase (encoded by theacdS gene) catalyses the conversion of ACC, theimmediate plant precursor for ethylene, into NH3 and-ketobutyrate and is widely distributed in soil fungiand bacteria, especially the Proteobacteria (Blaha et al.2006; Prigent-Combaret et al. 2008). Whilst theecological significance of ACC-deaminase in soilmicroorganisms is largely unknown, in plant-beneficialbacteria it may serve to diminish the amount of ACCavailable for production of ethylene (Glick et al.2007a). Since ethylene inhibits growth and elongationof root, this may lead to enhanced root systemdevelopment (Glick et al. 2007a). Indeed, this modelhas been validated by analysis of the root growthpromoting effect of a Pseudomonas putida PGPR,where acdS was inactivated (Glick et al. 1994; Li et al.2000). In the case of Azospirillum, complementation ofAcdS- strains with an acdS gene from P. putidaenhanced the plant-beneficial effects of these PGPRon both tomato (Lycopersicon esculentum Mill.) andcanola (Holguin and Glick 2001; Holguin and Glick2003). Similar results were obtained following theintroduction of the Pseudomonas acdS gene into AcdS-

Escherichia coli, Agrobacterium tumefaciens andbiocontrol strains of Pseudomonas spp., where expres-sion of ACC deaminase promoted root elongation,inhibition of crown gall development and improvedprotection against phytopathogens, respectively (Hao etal. 2007; Shah et al. 1998; Wang et al. 2000). From anumber of studies it appears that the growth promotioneffect of ACC deaminase in rhizobacteria is mosteffective in stress environments such as in flooded,heavy-metal contaminated or saline soils (Cheng et al.2007; Farwell et al. 2007) and in response tophytopathogens (Wang et al. 2000).

Cross-talk between plant-growth promoting pathwaysin plants

It is evident that plant regulatory molecules (e.g.auxin, ethylene, NO, gibberellin etc) do not act alonebut interact with one another in a variety of complexways (Fu and Harberd 2003; Glick et al. 2007a;Lamattina et al. 2003). Moreover, it is clear that thePGPR effect occurs as a result of a combination ofdifferent mechanisms (additive hypothesis). For ex-ample, a model has been proposed by Glick et al.2007a to describe cross-talk between auxin andethylene in both PGPR and plants. In response toroot exudates containing tryptophan, PGPR produceIAA that can be taken up by plant cells. Besides thedirect effect of IAA on plant cell proliferation andelongation, it also induces the synthesis of ACCsynthase in plants (Abel et al. 1995) and thereby theproduction of ethylene. A negative feedback loop,involving inhibition by ethylene of the transcriptionof auxin response factors, would lead in fine to a slowdown of ACC synthase activity and decrease of ACCand ethylene biosynthesis (Glick et al. 2007a). Otherclose interactions between IAA and ethylene path-ways have been recently reported. It appears thatethylene triggers the accumulation of auxin in the rootapex (Stepanova et al. 2005) and that the transport ofauxin from the apex to the elongation zone of roots isrequired for ethylene to inhibit root growth (Swarupet al. 2007). Overall, the molecular mechanisms bywhich ethylene and auxin interact to competitivelyregulate root development remain largely unknownand future prospects will aim to clarify their respec-tive contribution. AcdS- and IAA-producing PGPRmight promote root growth by both a lowering ofplant ethylene production and ethylene-dependentsignalling pathways, and through an increase in anethylene-independent manner, of the content of IAAin roots.

Facilitating plant iron and vitamins absorption

In addition to phytohormones, PGPR may influence thegrowth of plant roots through the production of side-rophores and vitamins. Roots of strategy II type plants(e.g. the Graminaceae; Marschner 1995; Robin et al.2008) secrete phytosiderophores (Fe-chelators) whichbind Fe3+ and maintain its concentration in soilsolution (see Lemanceau et al. 2009). At the root

Plant Soil (2009) 321:305339 323

surface chelated Fe3+ can then be directly taken upas a phytosiderophore-Fe complex (strategy II; seeLemanceau et al. 2009), whereas in non-graminaceousspecies (strategy I type plants; Marschner 1995; Robinet al. 2008) Fe3+ must first be reduced to Fe2+ prior tobeing absorbed by the plant (Lemanceau et al. 2009).Rhizobacteria (and fungi) also produce siderophoresand it has been shown that plants can absorb bacterial-Fe3+ complexes, which includes strategy I speciespossibly by endocytosis (Bar-Ness et al. 1991; Vansuytet al. 2007). The capture of these bacterial complexesby plants may play a significant role in nutrition andgrowth, especially in alkaline and calcareous soilswhere Fe availability is low (Bar-Ness et al. 1991;Masalha et al. 2000). Moreover, this mechanism isinvolved in biocontrol activities of PGPR and has beenlinked to competitive effects with phytopathogens andother detrimental rhizosphere microorganisms (Duijffet al. 1993; Longxian et al. 2005; Robin et al. 2008).

Plants under optimal growing conditions synthesizevitamins but when grown in stressed environments,vitamin-producing rhizobacteria may stimulate plantgrowth and yield. In particular, production of vitaminsof the B group (e.g. thiamine, biotin, riboflavine, niacin)has been documented in some Azospirillum, Azotobac-ter, Pseudomonas fluorescens and Rhizobium strains(Marek-Kozaczuk and Skorupska 2001; Revillas et al.2000; Rodelas et al. 1993; Sierra et al. 1999). There isevidence that exogenous supply of B-group vitamins toplants favours root development (Mozafar and Oertli1992), but there is presently no direct evidence (e.g.using mutants) that PGPR can stimulate plant growththrough this mechanism (Marek-Kozaczuk andSkorupska 2001), although further work is warranted.

Interactions between plant growth promotingmicroorganisms

Mycorrhizal associations

Microbial populations in the rhizosphere, includingknown PGPR, can either interfere with or benefit theformation and function of mycorrhizal symbioses(Gryndler 2000). A typical beneficial effect is thatexerted by mycorrhizal-helper-bacteria (MHB)which stimulate mycelial growth and/or enhancemycorrhizal formation (Garbaye 1994). This appliesboth to ectomycorrhizal fungi (Frey-Klett et al. 2005)and to AM fungi (Barea et al. 2005b; Johansson et al.

2004) and involves a range of bacterial species,commonly including Bacillus and Pseudomonas.Responses to MHB are associated with both theproduction of compounds that increase root cellpermeability and rates of root exudation, which eitherstimulate AM fungal mycelia in the rhizosphere orfacilitate root penetration by the fungus, and theproduction of phytohormones that influence AMestablishment (Barea et al. 2005b). Specific rhizobac-teria are also known to affect the pre-symbiotic stagesof AM development, such as spore germination andrate of mycelial growth (Barea et al. 2005b). Recently,Frey-Klett et al. (2007) revisited the significance ofMHB and differentiated the effects based on eitherAM formation or AM function, including nutrientmobilization, N2 fixation and protection of plantsagainst root pathogens.

Given that the external mycelium of mycorrhizasact as a link between roots and the surrounding soil,the fungus can also synergistically interact with soilmicroorganisms that mobilize soil P, through eithersolubilization or mineralization (Azcon et al. 1976;Barea 1991; Barea et al. 2005a; Kucey 1987; Tarafdarand Marschner 1995). Such interactions have beeninvestigated with 32P-based methodologies usingreactive rock phosphate in a non-acidic soil (Toro etal. 1997) and in an experiment conducted by Barea etal. (2002) using various treatments that included i)AM inoculation, ii) PSB inoculation, iii) AM plusPSB dual inoculation and iv) non-inoculated controlsin a soil containing natural populations of both AMfungi and PSB (Fig. 5). Soils were either un-amended(without P application) or fertilized with rock phos-phate. Both rock phosphate addition and microbialinoculation improved biomass production and Paccumulation in plants, with dual inoculation beingthe most effective (Barea et al. 2002; Fig. 5).Independent of rock phosphate addition, AM-inoculatedplants showed lower specific activity for 32P thancompared to non-AM inoculated controls, particularlywhen inoculated with PSB, suggesting that the PSBwere effective in releasing P from sparingly solublesources either directly from the soil or from added rockphosphate. Other studies have similarly showed apositive interaction between ectomycorrhizal fungiand the presence of bacterial isolates that showpotential for the weathering of soil minerals and itsassociated release of nutrients for plant uptake (Uroz etal. 2007). On such evidence it suggested that mycor-

324 Plant Soil (2009) 321:305339

rhizas are highly effective for improving the capture ofmobilized nutrients in soil, especially in relation to themobilization and capture of soil P.

Microbial interactions and enhanced nitrogen fixation

Interaction between mycorrhizal fungi, PGPR anddiazotrophs, including both rhizobia and associativeN-fixers, has also received considerable attention. In

legumes it is evident that AM can improve bothnodulation and N2 fixation within nodules (Barea etal. 2005a; Barea et al. 2005b). Co-inoculation ofRhizobium sp. with AM fungi gave greater growthpromotion in lucerne (alfalfa) and pea than inocula-tion of either symbiont alone (Hflich et al. 1994).The physiological and biochemical basis of AMfungal x Rhizobium interactions in improving legumeproductivity suggests that the main effect of AM inenhancing nodule activity is through a generalizedstimulation of host nutrition, but specific hormonaleffects on root and nodule development may alsooccur (Barea et al. 2005a).

Several re