The role of motility and cell surface hydrophobicity in trans-port and dispersal of Pseudomonas fluorescens strainsLAM1-hydrophilic, LAM2-hydrophobic and LAMNM (non-motile mutant of LAM2) under different soil conditions wasstudied. Maximum adhesion was recorded for LAM2 in clayloam (70%), followed by sandy loam (68%) and sandy soil(40%). Vertical migration of P. fluorescens isolates in soilswas recorded at 5 and 25 cm flow of wafer or M. phaseolinaexudate. In all the treatments, LAM1 exhibited maximummigration followed, by LAM2 and LAMNM. The rate ofmigration of such isolates was lowered in water irrigated soilscompared to those irrigated with M. phaseolina exudate. Insandy soil, cells of LAM1 migrated up to 13 cm in compari-son to LAM2 (11 cm) and LANNM (9 cm) at 5 cm flow offungal exudate. Population of LAM1, LAM2 and LAMNM

was 5.7, 5.68 and 5.61 log cfu g–1 soil at 1 cm depth, but itdecreased to 2.56, 2.21 and 1.99 log cfu during migration upto 11 cm in sandy soil at 5 cm flow of fungal exudate. Greatermotility was observed in sandy soil irrigated with water orfungal exudate, followed by sandy loam and clay loam. Ingeneral, filtration coefficient (λ) of P. fluorescens was higherin soils irrigated with 5 cm of water or exudate than with25 cm of irrigation. The horizontal movement of P. fluores-cens strains in sandy soil adjusted at different ψm showedmarked reduction with decrease in ψm. The non-motileLANNM did not show chemotactic response and migrated upto a maximum of 3 mm in saturated soils (0 kPa). After 96 h,LAM1 and LAM2 migrated upto 35 and 29 mm respectivelyin sandy soil. Motile isolates had significantly greater coloni-zation of M. phaseolina sclerotia over the non-motile mutant.

Key words: Macrophomina phaseolina – Pseudomonasfluorescens – chemotaxis – motility.

Introduction

Extensive work has been done to understand the bio-chemical and molecular aspects of bacterial motilityand chemotaxis (Stock and Surette 1996; Armitage andSchmitt 1997; Parent and Devreotes 1999). However,ecological significance of antagonistic motile soil bac-teria in transport through different soil types by attrac-tion and dispersion, has been emphasized invariably asit is affected by soil moisture, pore size, the rate ofwater percolation (Roughley and Worral 1984; Howieet al. 1987; Briettenbeck et al. 1988; Hyusman andVerstraete 1993; Toyota et al. 2000), filtration, adsorp-tion, desorption, growth, decay, sedimentation andchemotaxis, is not well understood (Bowers and Parke1993; Vande Broek 1998). Bacterial chemotaxis andmotility have a pivoted role in bioremediation on bio-facilitated transport of pollutants (Schafer et al. 1998),root nodulation, rhizosphere competence (Catlow et al.1990), colonization of roots (Bashan and Holugin1994; Toyota and lkeda 1997; Turnbull 2001), inter-actions with eukaryotic microbes (Gupta et al. 1991),and all these collectively attach the ecological signif-icance to bacterial motility (Broek et al. 1998).

In a complex soil environment, bacterial motility andchemotaxis seem to play important role in movementtowards the food resource, exudates from plant roots,eukaryotic organisms or nutrient release from decaying/dead plant parts (Scher et al. 1985; Bashan 1986; Aroraand Gupta 1993; Singh and Arora 2001). Though themode of bacterial motility in soils and around plantroots has been investigated by many (Howie et al. 1987,Briettenbeck et al. 1988; Parke et al. 1988; Boelens et al. 1994; Toyota and lkeda 1997; Turnbull 2001),

0944-5013/02/157/02-139 $15.00/0 Microbiol. Res. 157 (2002) 2 139

Microbiol. Res. (2002) 157, 139–148http://www.urbanfischer.de/journals/microbiolres

Horizontal and vertical movement of Pseudomonas fluorescenstoward exudate of Macrophomina phaseolina in soil : influence ofmotility and soil properties

Tanuja Singh, Alok K. Srivastava, Dilip K. Arora

Laboratory of Applied Mycology, Center of Advanced Study in Botany, Banaras Hindu University,Varanasi 221005, India

Accepted: January 17, 2002

Abstract

Corresponding author: D. K. Arorae-mail : dkarora@banaras.ernet.in

very little is known regarding the significance of bac-terial motility towards fungal exudate or propagules/mycelium in soil. Bacterial motility and chemotaxistoward the fungal propagules involves the most criticalstep in initiation of the fungal-bacterial interaction orthe subsequent colonization of spores (Hyakumachi andArora 1997). Earlier studies demonstrated that fungalpropagules/mycelia are also important in bacterialmovement/growth in soil as hyphae providing themigration pathway (Arora et al. 1983; Gupta et al.1991; Lim and Lockwood 1988; Arora and Gupta1993; Singh and Arora 2001). Attempts in this directionrevealed that attraction of bacteria and their growtharound fungal propagules/mycelia is primarily due tocompounds such as amino acids, carbohydrates, organicacids, polyols, amino sugars etc. exuded from the fungi(Arora and Gupta 1993; Singh and Arora 2001).

In the recent past, benefits of root colonizing andantagonistic fluorescent Pseudomonas spp. receivedmuch attention regarding their role (s) in increasingplant growth/productivity, and biological control ofsoilborne plant pathogens (Kloepper 1992; Lam andGaffeny 1993), although the observations on the bio-logical control of soilborne plant pathogens by P. fluo-rescens are frequently inconsistent owing partly to thelack of sufficient knowledge/strategies prerequisite topromoting colonization of roots or the deleteriouspathogenic fungi. One of such poorly understood traitsin the mechanism is lack of sufficient knowledge aboutthe motility behaviour and chemotactic response of suchbiocontrol bacteria in the soil environment and theirpossible roles in plant-bacterial and fungal-bacterialinteractions. Earlier studies on fluorescent pseudo-monads demonstrated that such bacteria can also becarried passively by water flowing through saturatedsoil (Roughley and Worral 1984) or move verticallythrough percolating water or attracted chemotacticallytowards roots on fungal propagules (Arora and Gupta1993; Bashan and Holugin 1994), indicating the role ofbacterial motility in plant-microbe and fungal-bacteriainteraction.

In the present investigation, two wild-type motileantagonistic P. fluorescens isolates (LAM1-hydrophil-ic ; LAM2-hydrophobic) and non-motile Tn5 mutant ofLAM2 were used to evaluate the vertical migrationthrough percolating water/fungal exudate in three soiltypes (sandy, sandy loam and clay loam). The horizon-tal movement and chemotactic response of P. fluores-cens isolates toward exudate of sclerotia of Macropho-mina phaseolina (a soilborne fungal pathogen causingcharcoal/root rot to 400 economic important crop plants)at different soil water matric potential was evaluated.The relative importance of motility/chemotaxis in suchisolates in their dispersal and subsequent colonizationof M. phaseolina sclerotia was also investigated.

Materials and methods

Organisms and growth conditions. Isolates of P. fluo-rescens from the rhizoshere of a tomato field (Gupta1995) were stored at –20°C in a 50% glycerolLuria-Burtani (LB) medium and routinely cultured inKing’s B medium (KB). The isolates were screened forcell surface hydrophobicity by BATH assay (Rosenberget al. 1980). Strains LAM1 and LAM2 were hydro-philic and hydrophobic, respectively (Fig. 3). Rifampi-cin marked strains of P. fluorescens LAM1 and LAM2were obtained by subculturing on KB supplementedwith rifampicin (Sigma, 25–250 µm ml–1). The re-sistant strains compared the parent with no evidence forreversion of marked strains up to more than 20 sub-cultures. Non-motile Tn5 mutant of P. fluorescensLAM2 was developed by transposan mutagenesis(Simon et al. 1983). In brief, it involved Tn5 insertionwith suicidal vector system of E. coli SM (Kan+) ob-tained from IACR, Rothamsted, U. K. Biparentalmatings were conducted between E. coli SM (donor)and rifampicin-resistant hydrophobic strain LAM2(recipient) from cultures grown overnight with appro-priate antibiotics. Donor (8 ml) and recipient (4 ml)cells mixed, harvested by centrifugation, resuspendedin LB and spotted onto HAWP membrane (0.45 µm,Millipore) on LB agar. After growth for 18–24 h(28–30°C) the bacteria were resuspended in 10 ml ofLB broth and trans-conjugants selected by dilutionplating on KB media supplemented with rifampicin andkanamycin (250 µg ml–1 each). The motility of trans-conjugants was tested in chemotaxis buffer (10 mM,KPB buffer containing 0.1 mM Na-EDTA, 1 mMMgSO4 and 0.2 mM CaCl2; pH 7). The non-motilemutant were grown to tryptone semi-solid medium for5 transfers and examined under phase contrast micro-scope to ensure the stability of non-motile character-istics. Wild rifampicin-resistant LAMI, LAM2 and Tn5non-motile mutant of LAM2 (LAMNM) were usedthroughout the study.

Macrophomina phaseolina Tassi (Goid), causingcharcoal rot of chickpea, was cultured on syntheticmedium (SM; Srivastava et al. 1996). The sclerotiawere obtained by brushing the 15 d old cultures grownon cellophane membrane spread over SM medium. Tocollect the fungal exudate, freshly harvested sclerotia(ca. 6 × 108 sclerotia ml–1) were suspended in 50 mMphosphate buffer (pH 7) and incubated for 48 h at 30°C.Exudate was then collected by passing through a sterilemembrane filter (Millipore; 0.45 µm) and filtrate wasused for saturation of test soils or irrigation treatment ofsoil columns.

Soils. Sandy, sandy loam and clay loams soils werecollected from different agricultural fields nearby

140 Microbiol. Res. 157 (2002) 2

Varanasi. The organic content of sandy, sandy loam andclay loam was 3.25, 4.2 and 2.67%, respectively. Themoisture characteristics of repacked soil samples wasevaluated in terms of the amount of water held in thesoil at different ψm. The effective pore neck diameterwas calculated by the formula: pF = Log 0. 15 – Log r ;where r = radius of the capillary pore (Vargas andHattori 1986).

Adhesion of P. fluorescens isolates to soil. Adhesion ofLAM1, LAM2 and LAMNM to sandy, sandy loam andclay loam soils was compared by modified method ofBrittenbeck et al. (1988). In brief, exponential phasebacteria were washed and resuspended in 150 mMNaCl (109 cells ml–1) and non-sterile air dried soil (1 g)was mixed with 10 ml of cell suspension in a sterilizedglass tube, vortexed for 1 min and shaken for 20 min toensure adequate contact between soil particles andcells. The suspension was centrifuged (100 g; 30 s) tosettle the soil particle with adhered bacteria and numberof cells in the supernatant were determined by dilutionplating on KB media amended with appropriate anti-biotics. The aliquot of the suspension was sampledapproximately from the middle of the tube. The ad-hesion was determined by the differences between thenumber of bacteria in solution before and after theadhesion test and expressed as percentage of the initialnumber in solution.

Transport of P. fluorescens through vertical soilcolumn. Polyvinyl chloride (PVC) plastic columns(25 cm long; 2.15 dia.), having one cm circular hole onside walls at every 2 cm, was packed uniformly withone dried sieved soil up to 20 cm to give an averagebulk density of 1.28, 1.33 and 1.38 mg M–3 for sandy,sandy loam and clay loam soil, respectively. Beforepacking, the column holes were sealed by cellophanetape. The soil was held at the bottom by cheese clothand nylon gauze. The column was saturated with wateror exudates of M. phaseolina sclerotia applied from the column top at a flow rate of 2 ml h–1 in order toexpel the air bubbles. After saturation, one ml cellsuspension (ca. 108 cells ml–1) of P. fluorescens isolateswas placed on the top of the column. A 2.5 cm casinglayer of acid washed sterile coarse sand was added to the top of the soil column in order to evenly distri-bute the irrigated water or fungal exudate. The flow rate was controlled by peristaltic pump. The soilsamples were taken through the column holes. To avoidcontamination, sampling started through holes frombottom to top and bacterial population in each samplewas determined by dilution plating on KB mediasupplemented with rifampicin or rifampicin+ kana-mycin (250 µg ml–1). Transport of P. fluorescens cellswas calculated: Log N = Log No – λ × 1; N = numberof bacteria (cfu g–1) at depth 1; No = initial number of

bacteria at the surface; λ = filtration coefficient ;1 = depth under the soil surface. The filtration coeffi-cient (λ) is an index of bacterial rate of transport. Thedepth where number of bacteria is reduced by 90%(D90) was also calculated (Hyusman and Verstraete1993).

Horizontal movement of P. fluoreseens in soil. Thehorizontal movement of P. fluorescens isolates towardexudate of M. phaseolina sclerotia and 10 µM glucosesolution (a known chemoattractant of P. fluorescens ;Arora and Gupta 1993) at different ψm in the soil wasstudied by using a modified version of apparatus (Aroraand Gupta 1993; Fig.1). In brief, a 500 ml Büchner fun-nel (100 mm diam), fitted with sintered glass plate offine porosity (0.45 µm), was used as tensiometer. Thelower end of funnel was sealed by rubber stopper con-nected with glass tubes travelling in the funnel. Theopen end of one tube was connected with a watercolumn, whereas the other tube was fitted with a stop-per to remove air bubbles. Sandy soil (100 g ) wasplaced in the funnel leaving a 5 mm central well andsaturated by water or exudate of M. phaseolina sclero-tia or 10 µM glucose solution. Soil was held at differentψm by adjusting the height of water column. Cellsuspension (1 ml; 108 cells ml–1) of LAM1, LAM2 orLAMNM was placed in the central well (Fig. 1). The ver-tical spikes of hollow glass rod (i.d. = 1 mm) was in-serted (ca. 10 mm deep) in the soil. After 24–96 h, thespikes were removed. Spikes containing soil wereimmersed in KPB (10 ml), shaken vigorously to removeadhered soil and populations of P. fluorescens wereestimated by dilution plating of the suspension on KB media amended with appropriate antibiotics(250 µg ml–1).

Colonization via chemotaxis. The role of motility/chemotaxis in colonization of M. phaseolina sclerotiawas studied in the tensiometer. M. phaseolina sclerotia(ca. 1 × 107), lodged on water agar blocks (HiMedia;100 mm long × 2 mm dia.), were buried in the soil heldin the Büchner funnel at a distance of 20 mm from cen-tral well containing LAM1, LAM2 or LAMNM cells(100 µl ; ca. 0.7–0.8 × 108 cells). In control treatments,blocks lacked sclerotia. The setup was incubated atdifferent soil ψm for 15 days. The agar blocks con-taining sclerotia were removed and colonizing popula-tion of P. fluorescens cells determined by dilutionplating on antibiotics containing KB medium.

Colonization of sclerotia mixed with P. fluorescens iso-lates. Colonization of P. fluorescens isolates in soil wasalso evaluated. In a Petri dish, unsterilized sandy soil(25 g) was mixed separately with cell suspensions (107 cells g–1 soil) of P. fluorescens isolates and in-cubated for 15 h (28 ± 2°C) to establish the bacterial

Microbiol. Res. 157 (2002) 2 141

population. The sclerotia (ca. 105) were dislodged on anylon net (80 µm pore size) and buried 2–3mm deep inthe soil. The plates were incubated in a moist chamberat different ψm maintained by adding predeterminedamount of water after 48 h to each plate. After 15 d ofincubation, infected sclerotia were removed, homogen-ized in 10 ml of sterilized PBS and colonizing popula-tion determined by dilution plating on KB supple-mented with appropriate antibiotics.

All experiments were repeated at least twice to esta-blish reproducibility. Bacterial populations in verticaland horizontal movement of soil and colonization ofsclerotia were transformed as log10 and values expres-sed log10 cfu g–1 soil. Data are subjected to standarddeviation analysis.

Results

Moisture content and pore-neck diameter of the soilwas recorded at different soil water ψm (Table 1). Ingeneral, the pore-neck diameter was highest in sandysoil, followed by sandy loam and clay loam. At –5 kPa,the pore-neck diameter was 190, 150 and 100 µm with12, 19 and 48 g H2O per 100 g of sandy, sandy loam andclay loam soil, respectively. There was a positive rela-tionship (r = 0.89– 0.92; P = 0.01) between the mois-ture content and soil pore-neck diameter (Table 1).

Hydrophobicity and adhesion of P. fluorescens to soil

Hydrophobicity of isolates LAM1, LAM2 and LAMNM

was 6.2, 38.8 and 30.2, respectively (Fig. 2). Maximumadhesion of the cells was recorded for LAM2 in clayloam (70.0%), followed by sandy loam (68.0%) andsandy soil (40.0%). The non-motile mutant LAMNM

exhibited a more or less similar adhesion as its parent,whereas for the less hydrophobic LAM1 it was least(Fig. 2).

Vertical migration of P. fluorescens in repacked soilcolumn

Vertical migration of P. fluorescens isolates in re-packed soil column was recorded at 0, 5 and 25 cm flowrate of water or M. phaseolina exudate (Fig. 3). In allthe treatments, maximum migration was observed forLAM1, followed by LAM2 and LAMNM. For instance,LAM1 moved up to 9 and 5 cm, respectively in thesandy and clay loam soil columns irrigated with 5 cmwater flow. However, LAM2 and LAMNM showedmaximum migration up to 7 and 5 cm, respectively, in5 cm water irrigated sandy soil. The cell migration wasgreater in the soils irrigated with fungal exudates thanthe water irrigated one (Fig. 3). For example, after irri-gation of sandy soil with fungal exudate at a flow rateof 25 cm, LAM1 and LAM2 migrated up to 19 and13 cm, respectively whereas in similar columns irri-gated with water, it moved up to only 17 and 11 cm. Ingeneral, the cell number declined logarithmically with

142 Microbiol. Res. 157 (2002) 2

Fig. 1. Schematic line diagram of apparatus used to study thehorizontal movement of P. fluorescens at different soil watermatric potential (modified from Arora and Gupta, 1993)

Table 1. Pore-neck diameter and moisture content of sandy, sandy loam and clay loam soil at different soil water matricpotential

Soil Matric Potential (kPa) Pore-neck diameter (µm) Moisture content (g H2O 100–1 g soil)

Sandy Sandy loam Clay loam Sandy Sandy loam Clay loam

–5 190 150 100 12 19 48–10 48 35 30 9 16 44–15 29 21 20 7 15 41–30 16 12 10 6 13 40

increase in soil depth. Population of LAM1, LAM2 andLANNM was 5.7, 5.68 and 5.61 log cfu g–1 soil at 1 cmdepth which decreased to 2.56, 2.21 and 1.99 log cfuduring transport up to 11 cm in sandy soil at 5 cm flowof fungal exudate (Fig. 3). A marked increase in migra-tion and higher population was recorded when flow rateincreased from 5 to 25 cm. For example at 5 cm flow offungal exudate LAM1 migrated up to 13, 11 and 9 cmin sandy, sandy loam and clay loam soil, whereas, at25 cm flow rate of exudates the migration was 19, 15and 11 cm, respectively in similar soils. In all the treat-ments, the migrating population was significantly lessin sandy loam and clay loam compared to sandy soilirrigated with water or fungal exudate.

Filtration coefficient (λ) was higher in soils irrigatedwith 5 cm of water or fungal exudate in comparison to25 cm of irrigation (Table 2). For example, λ values ofLAM2 was 0.78, 1.34 and 1.50 at 5 cm and 0.43, 0.63

Microbiol. Res. 157 (2002) 2 143

Fig. 2. Adhesion (bar) and hydrophobicity (line) of differentisolates of P. fluorescens (LAM1, LAM2 and LAMNM) insandy (�), sandy loam ( ) and clay loam (�) soil. Data aremeans of three replicates.

Fig. 3. Vertical migration of P. fluorescens isolates in sandy, sandy loam and clay loam soils saturated or irrigated with wateror exudate of M. phaseolina sclerotia. Symbols �, �, �, � and �, � represents LAM1, LAM2 and LAMNM, respectively.Filled and blank symbols represent water irrigated and M. phaseolina exudate irrigated soils, respectively. Data are mean ofthree replicates. 1 = average maximal S. D.

and 0.76 at 25 cm water irrigated sandy, sandy loam andclay loam soil, respectively. A slight decrease in λvalues was observed when soils were irrigated with fun-gal exudate (Table 2). The D90 values of the isolatesincreased significantly when rate of water flow wasenhanced from 5 to 25 cm. For example, the D90 valueof LAM1, LAM2 and LANNM was 2.12, 1.28, and 1.7at 5 cm water flow which increased to 4.3, 2.33 and 3.0,respectively at 25 cm water flow in sandy soil. In gene-ral, the D90 values of the isolates inereased by 4–7%when soil was irrigated with M. phaseolina exudatecompared to water (Table 2). The least D90 values wereobtained for LANNM in clay loam soil, followed bysandy loam at both the irrigation levels.

Horizontal movement of P. fluorescens

The horizontal movement of P. fluorescens isolates wasstudied in sandy soil at different ψm (Fig. 4). The soilψm decrease coincided with decrease in the rate ofmigration. For instance after 96 h of incubation LAM1and LAM2 migrated up to 15 and 11 mm at 0 kPa; 13and 9 mm at –5 kpa; and 11 and 6 mm at –10 kPa,respectively. Chemotactic response of P. fluorescensisolates in the presence of attractants, i.e. M. phaseolinaexudate or glucose, was also evaluated at 0 kPa ψm(Fig. 4). The motile isolates LAM1 and LAM2 showed increased migration in soils saturated withfungal exudate. After 96 h, the LAM1 population wasca. 1.68 log cfu g–1 soil at 35 mm distance from thecentral well, whereas LAM2 migrated up to 29 mmwith population of ca. 1.3 log cfu g–1 soil. Similarly, therate of migration also increased in soils saturated withglucose. Non-motile LAMNM did not show significantchemotactic response and migrated maximum up to5 mm in glucose or fungal exudate-saturated soil (Fig. 4).

Colonization of P. fluorescens cells to sclerotia of M. phaseolina

Motile and hydrophobic P. fluorescens isolates showedsignificantly greater colonization of M. phaseolinasclerotia via chemotaxis in comparison to non-motileisolate at 0, –5, –10, –15 and –30 kPa Ψm (Fig. 5). Forinstance, the colonizing population of LAM2 was2.56 log cfu, followed by LAMI (2.22 log cfu) andLAMNM (0.52 log cfu) at 0 kPa. LAM1 and LAM2colonized the sclerotia up to –15 kPa, whereas coloni-zation by LAMNM was recorded only up to –5 kPa (Fig. 5A). The colonizing population was higher fortreatments where bacteria were mixed in soil and werein physical contact with the sclerotia. Maximum colo-nization was recorded for LAM2 (3.9 log cfu 50–1

sclerotia), followed by LAM1 (3.6) and LAMNM (3.5)

144 Microbiol. Res. 157 (2002) 2

Tabl

e2.

Filtr

atio

n C

oeff

icie

nt (

λ) a

nd D

90va

lue

of P

seud

omon

as f

luor

esce

nsis

olat

es m

igra

ted

thro

ugh

vert

ical

soi

l col

umns

P. f

luor

es-

Filtr

atio

n C

oeff

icie

nt (

λ)D

90va

lue

(cm

)ce

nsst

rain

sSa

ndy

Sand

y lo

amC

lay

loam

Sand

ySa

ndy

loam

Cla

y lo

am

AB

AB

AB

AB

AB

AB

Irri

gate

d w

ith

wat

er

LA

M1

0.47

±0.

040.

23±

0.02

0.72

±0.

080.

33±

0.03

0.89

±0.

080.

36±

0.03

2.12

±0.

084.

3±

0.13

1.39

±0,

053.

03±

0.16

1.12

±0.

042.

77±

0.9

LA

M2

0.78

±0.

040.

43±

0.03

1.34

±0.

090.

63±

0.03

1.50

±0.

010.

76±

0.05

1.28

±0.

092.

33±

0.02

0.74

±0.

051.

58±

0.10

0.66

±0.

051.

31±

0.01

LA

MN

M0.

59±

0.30

0.33

±0.

011.

24±

0.05

0.52

±0.

021.

31±

0.02

0.54

±0.

021.

7±

0.61

3.0

±0.

090.

80±

0.06

1.92

±0.

20.

76±

0.05

1.85

±0.

01

Irri

gate

d w

ith

M. p

hase

olin

a ex

udat

e

LA

M1

0.39

±.0

10.

22±

0.06

0.80

±0.

010.

30±

0.01

0.88

±0.

060.

34±

0.01

2,56

±0.

064.

54±

0.03

1.25

±0.

063.

33±

0.06

1.14

±0.

032.

95±

0.07

LA

M2

0.74

±0.

020.

40±

0.02

1.30

±0.

010.

61±

0.01

1.43

±0.

020.

72±

0.02

1.35

±0,

762.

50±

0.09

0.76

±0.

031.

64±

0.06

0.70

±0.

011.

92±

0.06

LA

MN

M0.

54±

0.03

0.32

±0.

021.

18±

0.02

0.43

±0.

031.

28±

0.01

0.55

±0.

021.

85±

0.08

3.13

±0.

020.

85±

0.13

2.3

±0.

030.

78±

0.05

1.82

±0.

09

Aan

d B

=ir

riga

tion

of 5

cm a

nd 2

5cm

of

wat

er, r

espe

ctiv

ely.

Dat

a ar

e m

eans

of

3 re

plic

ates

±S.

D.

at 0 kPa. Soil moisture played a significant role in thecolonization of sclerotia. For example, LAM1 coloni-zed the sclerotia by 3.6, 3.3, 2.4, 2.0 and 1.2 log cfu at0, –5, –10, –15 and –30 kPa., respectively. A significantnegative correiation (P = 0.05; r = –0.93–0.95) wasobserved between colonizing population and soil waterΨm. The colonization ability of the isolates LAM1,LAM2 and LAMNM on sclerotia did not differ signifi-cantly at different ψm (Fig. 5B).

Discussion

The results suggest that both vertical and horizontalmigration of P. fluorescens isolates were higher in soils in the presence of the specific chemical stimulus(M. phaseolina exudate or glucose) suggesting thatmotile isolates are attracted to energy-rich compoundspresent in the exudate of M. phaseolina, e.g. glucose,amino acids, polyols, sugar, alcohol, amino sugars and

Microbiol. Res. 157 (2002) 2 145

Fig. 4. Horizontal migration of P. fluorescens isolates in sandy, sandy loam and clay loam soils held at different soil watermatric potential. Symbols �, � and � represents isolates LAM1, LAM2 and LANNM, respectively. Data are mean of threereplicates; | = average maximal S. D.

organic acids (Singh and Arora 2001), which facilitatedthe positive transport of bacteria along nutrient flow(Arora and Gupta 1993). Therefore, these isolates coulduse the chemotactic route for migration in soil. Otherfactors such as the rate of percolating water, adhesionand hydrophobic properties of bacteria also play im-portant role in vertical migration and horizontal move-ment of P. fluorescens in different soils. For example,significant (P = 0.05) increase in the number of trans-ported bacteria was recorded at 15 cm depth with 25 cm water irrigated soil packed columns compared to 5 cm (Fig. 3). Other workers (Briettenbeck et al.1988; Trevors et al. 1990; Boelens et al. 1994) alsoreported that vertical migration of bacteria with floatingor percolating water appears to be determining factorfor their distribution and colonization of the appro-priate ecological niche in soil. The movement of P. fluorescens was significantly (P = 0.05) high in soils with large continuous pores (sandy) compared to

narrow water channels in clay or sandy loam soil (Fig. 2). Vertical transport of P. fluorescens in soil waspositively correlated with the soil pore-neck diameter ;however, differences in porosity can not fully accountfor the observed variation in motility. These findingsare in agreement with Van Elsas et al. (1991) andHyusman and Verstraete (1992) who also reported thatmigration of bacteria reduces with high bulk density insoil. In comparison to repacked soil columns, hightransport of bacteria was recorded in undisturbed soils(Smith et al. 1985, Van Elsas et al. 1991) because in the undisturbed soil column, water moves throughcrack slits and macropores. However, experiments withthe repacked soil partially eliminated the interferencedue to irregularities of soil structure and the macro-pores and it can be used for a more precise evaluationof other factors viz., bacterial cell characteristics,nutrients etc. on bacterial movement (Hyusman andVerstraete 1992).

The exponential decrease in number of P. fluorescensisolates with the column depth was influenced by bac-terial adhesion to soil particles and filtration of cellsthrough soil matrix. In comparison to LAM1 (hydro-philic), the high adherence of hydrophobic LAM2 was observed for all the soils. This is in agreement withthe earlier reports that cell surface hydrophobicity was crucial to bacterial migration (Van Loosdrecht et al. 1987; Jana et al. 2000). Gannon et al. (1991)reported that the transfer of latex microspheres (similarin size to bacteria) was significantly influenced by surface characteristics than its size. Our results de-monstrate that filtration coefficient (λ) of bacteria could be a measure of hydrophobicity influencing themigration of bacteria (Table 2). De Flaun et al. (1990)also demonstrated that percentage of bacteria, leachingout of soil column is a measure of adhesion to the soil. In this study, the hydrophobic strain LAM1 ex-hibited maximum migration compared to hydrophilicLAM2; however, in clay loam soil, the migration ofLAM1 was also significantly low (Fig. 3) and this isprobably due to filtration of bacteria by soil matrix and absorption on soil particles as pore-neck diameterand bacterial cell size influenced migration in waterfacilitated transport in clay soil (Gannon et al.1991).

The role of chemotaxis and motility in distribution of P. fluorescens in soil was studied by measuring thehorizontal movement of bacteria in sandy soil saturatedby water or M. phaseolina exudates and adjusted atdifferent soil Ψm. Maximum horizontal migration wasobserved at 0 kPa that decreased with lowering in thematric potential (Fig. 4). No significant (P = 0.05) dif-ference between migration of LAM1 and LAM2 wasrecorded in soils saturated with fungal exudate. LAMNM

failed to migrate, although a few cells were detected

146 Microbiol. Res. 157 (2002) 2

Fig. 5. Colonization of M. phaseolina sclerotia in sandy soilby P. fluorescens isolates at different soil water matric poten-tial. A, colonization of P. fluorescens to M. phaseolina bychemotaxis ; B, colonization by P. fluorescens isolates mixedwith the sclerotia of M. phaseolina (see Materials andMethods for detail). Symbols �, and � represents isolatesLAM1, LAM2 and LAMMN, respectively; data are means of20 replicates.

up to 10 mm in soils saturated with exudate that mightbe due to sensory excitation of some chemoreceptorscausing increase in tumbling but not in directionalswimming (Fig. 4).

Bacterial migration toward fungal propagules is animportant step in the colonization of propagules(Hyakumachi and Arora 1998) that helps to initiateinteraction of specific bacteria with fungal propagules.Our earlier studies showed that P. fluorescens isolateswere attracted towards the exudates of different types offungal propagules containing various amount of ener-gy-yielding compounds (Arora and Gupta 1993). In thisstudy, the colonization of M. phaseolina sclerotia by P.fluorescens strains was recorded at different soil Ψm(Fig. 5). In soils mixed together with P. fluorescens andM. phaseolina sclerotia, insignificant differences incolonizing populations of LAM1, LAM2 and LAMNM

was recorded. This clearly indicates that while in closeproximity of attractants, the motility/chemotaxis maynot be significant for colonization of fungal propagules(Fig. 5B). Greater colonization of motile isolates wasrecorded when placed at a distance of 20 mm from fun-gal propagules thus elucidating the role of motility andchemotaxis in colonization. The colonization by P. flu-orescens was always greater in saturated soil. This may be because at 0 kPa Ψm higher rate of diffusion ofexudates may occur, and decrease in Ψm may alsodecrease the diffusion of solutes in soil (Griffin 1972;Filonow and Arora 1987). In addition as soil dries, thewater filled pores become narrower and thus inter-rupting the continuous water pathways necessary fordirectional motility and subsequent colonization offungal propagules (Arora et al. 1983; Arora and Gupta1993).

Overall, the results demonstrate that vertical trans-port of P. fluorescens through different soil typesdepends upon the physical properties of the latter e.g.,texture and pore size, that determine adsorption of bac-teria. The data strongly suggest that under continuouswater flow motility does not play a significant role invertical transport of P. fluorescens ; however, fungalexudate or chemoattractants like glucose facilitate suchmigration. Motility of P. fluorescens plays a significantrole in horizontal migration in the presence of a chemi-cal stimulus. Motility and chemotaxis are required forincreased colonization of M. phaseolina sclerotia in soilat a ψm sufficient enough to permit the bacterial move-ment and this could be significant in early recognitionsteps of fungal-bacterial interaction (Jana et al. 2000) insoil giving motile bacteria a competitive advantageover the non-motile ones. Colonization of pathogenicfungal propagules by antagonistic bacteria is also im-portant for weakening of propagules leading to loss ofviability and pathogenic aggressiveness (Hyakumachiand Arora 1998). Further studies are needed to evaluate

the significance of motility/chemotaxis of other antago-nistic/PGPR bacteria in soils at different environmentalconditions and its subsequent ecological role in thecolonization of fungal propagules in soil.

References

Armitage, J. P., Schmitt, R. (1997): Bacterial chemotaxis :variations on a theme: Rhodobaeter sphaeroides andSinorhizobium meliloti. Microbiol. 143, 3671–3682.

Arora, D. K., Filonow, A. B., Lockwood, J. L. (1983) :Bacterial chemotaxis to fungal propagules in vitro and insoil. Can. J. Microbiol. 29, 1104–1109.

Arora, D. K., Gupta, S. (1993): Effect of different environ-mental conditions on bacterial chemotaxis toward fungalspores. Can. J. Microbiol. 39, 922–931.

Bashan, Y. (1986): Migration of the rhizosphere bacteriaAzospirillum brasilense and Pseudomonas fluorescenstowards wheat roots in the soil. J. Gen. Microbiol. 132,3407–3414.

Bashan, Y., Holugin, G. (1994): Root-to-root travel of thebeneficial bacterium Azospirillum brasilense. Appl.Environ. Microbiol. 60, 2120–2131.

Boelens, J., Woestyne, M. V., Verstraete, W. (1994) :Ecological importance of motility for the plant growth-promoting rhizopseudomonas strain ANP 15. Soil Biol.Biochem. 26, 269–277.

Bowers, J. H., Parke, J. L. (1993): Colonization of Pea (Pisumsativum L.) taproots by Pseudomonas fluorescens: effect ofsoil temperature and bacterial motility. Soil Biol. Biochem.25, 1693–1701.

Briettenbeck, G. A., Yang, H., Dunigan, E. P. (1988) :Water-facilitated dispersal of inoculant Bradyrhizobiumjaponicum in soils. Biol. Fertil. Soils 7, 58–62.

Broek, A. V., Vanderleyden, J. (1995): The role of bacterialmotility, chemotaxis and attachment in bacterial-plantinteraction. Mol. Plant-Microbe Interact. 8, 800–8 10.

Broek, A. V., Lambrecht, M., Vanderleyden, J. (1998) :Bacterial chemotactic motility is important for the initiationof wheat root colonization by Azospirillum brasilense.Microbiol. 144, 2599–2606.

Catlow, H. Y., Glenn, A. R., Michael, J. D. (1990): The use ofTransposan-induced non-motile mutants in assessing thesignificance of motility of Rhizobium leguminosarum bio-var trifolii for movement in soils. Soil Biol. Biochem. 22,331–336.

DeFlaun, M. F., Marshall, B. M., Kulle, E. P., Levy, S. B.(1994): Tn5 insertion mutants of Pseudomonas fluorescensdefective in adhesion to soil seeds. Appl. Environ.Microbiol. 60, 2637–2642.

Filonow, A. B., Arora, D. K. (1987): Influence of soil matricpotential on 14C exudation from fungal propagules. Can. J.Bot. 65, 2084–2089.

Gannon, J. T., Manilal, V. B., Alexander, M. (1991) :Relationship between cell surface properties and transportof bacteria through soil. Appl. Environ. Microbiol. 57,190–193.

Griffin, D. M. (1992) : Ecology of soil fungi. SyracuseUniversity Press, Syracuse, N.Y. pp. 1–93.

Microbiol. Res. 157 (2002) 2 147

Gupta, S., Arora, D. K., Rai, B. (1991): Ecological aspects ofmicrobial chemotaxis. J. Sci. Res. Banaras HinduUniversity, 41, 21–39.

Gupta, S., Arora, D. K., Srivastava, A. K. (1995): Growth pro-motion of tomato plants by rhizobacteria and imposition ofenergy stress on Rhizoctonia solani. Soil Biol. Biochem.27, 1051–1058.

Howie, W. J., Cook, R. J., Weller, D. M. (1987): Effects of soilmatric potential and cell motility on wheat root coloniza-tion by fluorescent pseudomonads suppressive to take-all.Phytopathol. 77, 286–292.

Hyakumachi, M., Arora, D. K. (1998): The response of fungalpropagules to competitive energy stress in soil. In: (Eds.Csermely, P.) Stress of Life from Molecules to Man. Annalsof the New York Academy of Sciences, New York,152–162.

Hyusman, F., Verstraete, W. (1992): Effect of cell surfacecharacteristics on the adhesion of bacteria to soil particles.Biol. Fertil. Soils 16, 21–26.

Hyusman, F., Verstraete, W. (1993): Water-facilitated trans-port of bacteria in unsaturated soil columns: influence ofcell surface hydrophobicity and soil property. Soil Biol.Biochem 25, 83–90.

Jana, T. K., Srivastava, A. K., Csery, K., Arora, D. K. (2000):Influence of growth and environmental conditions on cellsurface hydrophobicity of Pseudomonas fluorescens innon-specific adhesion. Can. J. Microbiol. 46, 28–37.

Kloepper, J. W. (1992): Plant growth promoting rhizobacteriaas biological control agents. In: Soil Microbial Tech-nologies (Ed: Matting, B.) Marcel Dekker, New York,142–152.

Lam, S.T., Gaffney, T. D. (1993): Biological activities ofbacteria used in plant pathogen control. Biotechnology inplant disease control. Wiley-Liss, Inc., 291–319.

Lim, W. C., Lockwood, J. L. (1988): Chemotaxis to somephytopathogenic bacteria to fungal propagules in vitro andin soil. Can. J. Microbiol. 34, 196–199.

Parent, C., Devreotes, P. (1999): A cell’s sense of direction.Science, 284, 765–770.

Parke, J. L., Moen, R., Rovira, A. D., Bowen, G. D. (1986):Soil water flow affects the rhizosphere distribution of aseed-borne biological control agent, Pseudomonas fluores-cens. Soil Biol. Biochem. 18, 583–588.

Rosenberg, M., Gutnick, D., Rosenberg, E. (1980) :Adherence of bacteria to hydrocarbons: a simple methodfor measuring cell-surface hydrophobicity. FEMS Micro-biol. Lett. 9, 29–33.

Roughley, R. J., Worral, V. S. (1984): Vertical movement ofRhizobium trifolli in soil. Abstract. In Australian Institute of Agricultural Science (AIAS) Occasional Publication No. 12 (Eds. Kennedy, I. R. and Copeland, L.) The 7th

Australian Legume Nodulation Conference, University ofSydney.

Schafer, A., Ustohal, P., Harms, H., Stauffer, F., Dracos, T.,Zehnder, A. J. B. (1998): Transport of Bacteria in un-

saturated porous media. J. Contaminant Hydrol. 33,149–169.

Scher, F. M., Kloepper, J. W., Singleton, C. A. (1985) :Chemotaxis of fluorescent Pseudomonas spp. to soybeanseed exudates in vitro and soil. Can. J. Microbiol. 31,570–574.

Singh, T., Arora, D. K. (2001): Motility and chemotacticresponse of Pseudomonas fluoreseens toward attractantspresent in the exudate of Macrophomina phaseolina.Microbiol. Res. 156, 343–351.

Simon, R., Priefer, U., Puhier, A. (1983): A broad host rangemobilization system for in vivo genetic engineering: trans-posan mutagenesis in gram negative bacteria. BioTechnol.1, 784–789.

Smith, M. S., Thomas, G. W., White, R. E., Ritonga, D.(1985): Transport of Escherichia coli through intact anddisturbed soil columns. J. Environ. Qual. 14, 87–91.

Srivastava, A. K., Arora, D. K., Gupta, S., Pandey, R. R., Lee,M. W. (1996): Diversity of potential antagonists colonizingsclerotia of Macrophomina phaseolina in soil, and agglu-tination properties of some potential antagonists towardsclerotia. Biol. Fertil. Soils 22, 136–140.

Stock, J. B., Surette, M. G. (1996): Chemotaxis. In: Esche-richia coli and Salmonella typhimurium. Cellular andMolecular Biology (Eds. Neidhardt, F. C.) 1, 1103–1129.

Toyota, K., lkeda, K. (1997): Relative importance of motilityand antibiosis in the rhizoplane competence of a biocontrolagent Pseudomonas fluorescens MeIRC2Rif. Biol. Fertil.Soils 25, 416–420.

Toyota, K., Kimura, M., Kinoshita, T. (2000): Microbio-logical factors affecting the colonization of tomato roots by Ralstonia solanacearum YUlRify3lyx. Soil Sci. PlantNutri. 46, 643–653.

Trevors, J. T., Van Elsas, J. D., Van Overbeek, L. S., Starodub,M. E. (1990) : Transport of a genetically engineeredPseudomonas fluorescens strain through a soil microcosm.Appl. Environ. Microbiol. 56, 401–408.

Turnbull, G. A., Morgan, J. A. W., Whipps, J. M., Saunders,J. R. (2001): The role of motility in the in vitro attachmentof Pseudomonas putida PaW8 to wheat roots. FEMSMicrobiol. Ecol. 35, 57–65.

Van Elsas, J. D., Van Overbeek, L. S., Feldmann, A. M.,Dullemans, A. M., De Leeuw, O. (1991): Survival ofgenetically engineered Pseudomnas fluorescens in soil incompetition with the parent strain. FEMS Microbiol. Lett,77, 257–264.

Van Loosdrecht, M. C. M., Johannes, L., Willem, N., Gosse,S., Alexander, J. B. Z. (1987): Etectrophoretic mobility andhydrophobicity as a measure to predict the initial steps ofbacterial adhesion. Appl. Environ. Microbiol. 43,1898–1901.

Vargas, R., Hattori, T. (1986): Protozoan predation of bac-terial cells in aggregates. FEMS Microbiol. Ecol. 38,233–242.

148 Microbiol. Res. 157 (2002) 2