Selective Enhancement of the Fluorescent Pseudomonad Population After Amending the Recirculating...

ORIGINAL ARTICLE

Selective Enhancement of the Fluorescent PseudomonadPopulation After Amending the Recirculating NutrientSolution of Hydroponically Grown Plantswith a Nitrogen Stabilizer

D. Pagliaccia & D. Merhaut & M. C. Colao & M. Ruzzi &F. Saccardo & M. E. Stanghellini

Received: 26 February 2007 /Accepted: 5 February 2008 /Published online: 18 March 2008# Springer Science + Business Media, LLC 2008

Abstract Fluorescent pseudomonads have been associated,via diverse mechanisms, with suppression of root diseasecaused by numerous fungal and fungal-like pathogens.However, inconsistent performance in disease abatement,

after their employment, has been a problem. This has beenattributed, in part, to the inability of the biocontrolbacterium to maintain a critical threshold populationnecessary for sustained biocontrol activity. Our resultsindicate that a nitrogen stabilizer (N-Serve®, Dow Agro-sciences) selectively and significantly enhanced, by two tothree orders of magnitude, the resident population offluorescent pseudomonads in the amended (i.e., 25 μgml−1 nitrapyrin, the active ingredient) and recycled nutrientsolution used in the cultivation of hydroponically growngerbera and pepper plants. Pseudomonas putida wasconfirmed as the predominant bacterium selectively en-hanced. Terminal restriction fragment length polymorphism(T-RFLP) analysis of 16S rDNA suggested that N-Serve®selectively increased P. putida and reduced bacterialdiversity 72 h after application. In vitro tests revealed thatthe observed population increases of fluorescent pseudo-monads were preceded by an early growth suppression ofindigenous aerobic heterotrophic bacteria (AHB) popula-tion. Interestingly, the fluorescent pseudomonad populationdid not undergo this decrease, as shown in competitionassays. Xylene and 1,2,4-trimethylbenzene (i.e., the inertingredients in N-Serve®) were responsible for a significantpercentage of the fluorescent pseudomonad populationincrease. Furthermore, those increases were significantlyhigher when the active ingredient (i.e., nitrapyrin) and theinert ingredients were combined, which suggests a syner-gistic response. P. putida strains were screened for theability to produce antifungal compounds and for theantifungal activity against Pythium aphanidermatum andPhytophthora capsici. The results of this study suggest thepresence of diverse mechanisms with disease-suppressingpotential. This study demonstrates the possibility of using a

Microb Ecol (2008) 56:538–554DOI 10.1007/s00248-008-9373-z

D. Pagliaccia : F. SaccardoDepartment of Plant Production, University of Tuscia,01100, Viterbo, Italy

F. Saccardoe-mail: [email protected]

D. MerhautDepartment of Botany and Plant Sciences,University of California,Riverside, CA 92521, USAe-mail: [email protected]

M. E. StanghelliniDepartment of Plant Pathology, University of California,Riverside, CA 92521, USAe-mail: [email protected]

M. C. Colao :M. RuzziDepartment of Agrobiology and Agrochemistry,University of Tuscia,01100 Viterbo, Italy

M. C. Colaoe-mail: [email protected]

M. Ruzzie-mail: [email protected]

D. Pagliaccia (*)Department of Plant Pathology,University of California Riverside,Fawcett Laboratory 234, Riverside, CA 92521, USAe-mail: [email protected]

specific substrate to selectively enhance and maintaindesired populations of a natural-occurring bacterium suchas P. putida, a trait considered to have great potential inbiocontrol applications for plant protection.

Introduction

Greenhouse growers are facing strict governmentalregulations concerning the discharge of spent nutrientsolution to abate ground water pollution resulting fromnutrient salts and agricultural chemicals. In addition, theuse of recaptured irrigation water by the agricultural/horticultural industry is increasing due to pressure bygovernmental agencies to reduce water usage. Althoughthe use of recycled irrigation water will reduce waterusage and mitigate nutrient runoff from nursery produc-tion sites, serious grower concerns exist regarding thespread of phytopathogenic microorganisms via therecycled water. Management of these phytopathogenshas been attempted using various biological, cultural,chemical, and physical methods [42, 50, 59, 64]. Whilethese methods provide a limited measure of control forspecific pathogens, the search for new methods orapproaches is continuing.

During the course of our investigations on the controlof zoosporic pathogens in hydroponic cultural systems,we discovered that amending the recirculating nutrientsolution with N-Serve®, (a chemical which selectivelyinhibits denitrifying bacteria, i.e., Nitrosomonas spp.,Dow AgroSciences) resulted in suppression of root rotof pepper caused by Phytophthora capsici [40]. Inaddition, preliminary studies indicated that the populationof fluorescent pseudomonads appeared to be higher in N-Serve®-amended compared to non-amended treatments(authors, unpublished). Several fluorescent pseudomonadshave been associated, via diverse mechanisms, with thesuppression of root diseases caused by numerous fungaland fungal-like plant pathogens [18, 42, 58, 66]. Inaddition, some isolates are recognized as plant growthpromoters (PGP) and biopesticide/pollutant degraders [4,26, 47, 62, 65].

The commercial product of N-Serve®24 contains75.55% of inert ingredients (primary inert ingredients:23% 1,2,4-trimethylbenzene and 2% of xylene-mixedisomers), and 22.2% of active ingredient nitrapyrin [2-chloro-6-(trichloromethyl) pyridine], plus 2.25% relatedchlorinated pyridines. With respect to the role N-Serve®on the bacterial population, the degradation and utilizationof solvents (i.e., inert ingredients) such as xylene and 1,2,4-trimethylbenzene, as carbon sources by pseudomonads, hasbeen reported [29, 35, 53, 55]. No information is availableon the effect of nitrapyrin on the bacterial population and,

more specifically, on the pseudomonads in soilless systems.Nitrapyrin is a specific inhibitor of autotrophic ammoniaoxidation [16, 54], and ammonia is oxidized to hydroxyl-amine by ammonia monoxygenase (AMO). Campbell andAleem [5] reported that nitrapyrin inhibits the oxidation ofammonia to hydroxylamine by Nitrosomonas europaea.They suggested further that the inhibitor acts by chelatingthe copper components of the cytochrome oxidase involvedin ammonia oxidation and that this inhibition could bereversed by the addition of copper to the growth medium.Since then, conflicting reports have been published over theyears. While Hooper and Terry [19] confirmed Campbelland Aleem’s theory [5], Powell and Prosser [45] contra-dicted it by stating that copper enhances rather thanreverses the inhibition. In the 1990s, Vannelli and Hooper[63] showed that nitrapyrin is both an irreversible inhibitorof and a substrate for Nitrosomonas AMO. Nitrapyrin hasalso been reported as an alternative substrate for the AMOenzyme [20]. Nitrapyrin has further been reported as beingbactericidal effect, i.e., Nitrosomonas are not only tempo-rarily inhibited in their activity, but part of the population iskilled in treated soil [61]; while a large number of otherbacteria, actinomycetes, and fungi were found to beinsensitive to this molecule [16, 28, 54].

Although the specific mechanism of inhibition ofnitrapyrin remained unclear [19], its role as antimicrobialand potential metal chelator seems appealing, especially ina hydroponic system. This is because the formation ofcomplexes between soluble metals and nitrapyrin couldprovide essential metals for bacterial growth and may givea significant growth advantage to biocontrol bacterialspecies in the nutrient solution and the rhizosphere.

Biochelators, with different functions, are also pro-duced by microorganisms. Biochelators, such as thesiderophore pyoverdine, have specific roles in ironacquisition and transport by microorganisms [52], whileothers are responsible for the transport of some othermetals. Beside pyoverdine, most microorganisms produce avariety of biochelators with secondary functions such asredox activity, or molecules with antibiotic activity, orinduction of resistance against pathogens in plants [9, 10,32, 38].

Encouraged by preliminary results, and being aware ofthe potential chelating and antimicrobial properties ofnitrapyrin a.i., a series of experiments were conducted todetermine: (1) the effects of N-Serve® on the dynamics ofthe bacterial population in amended versus non-amendedhydroponic cultural systems; (2), the role of N-Serve®active ingredient and the ‘inert components’ on thefluorescent pseudomonad populations; (3), evaluate thechelating potential of nitrapyrin; and (4) assess the abilityof the test strain of P. putida to produce siderophore andother antifungal metabolites.

Selective Enhancement of the Fluorescent Pseudomonad Population 539

Materials and Methods

Greenhouse Studies

Hydroponic Cultural Systems

Experiments were conducted in closed recirculating hydro-ponic cultural systems in temperature-controlled greenhouses(20–32°C; Fig. 1). Two different host plants, gerbera(Gerbera jasemonii cv. Timo) and pepper (Capsicumannuum cv. Joe Parker) were used to assess the influenceof different plant species on the response of bacteria afterchemical amendments to the nutrient solution. A commer-cial nitrification inhibitor, (N-Serve® 24, Dow Agro-Sciences, Indianapolis) was used as the amendment tothe recirculating nutrient solution in greenhouse experi-ments. The commercial product contains 22.2% nitrapyrin,2.2% related chlorinated pyridines, and 75.55% inertingredients (xylene and 1,2,4-trimethylbenzene are theprimary inert ingredients).

Gerbera

The hydroponic recirculating system used in the gerberagreenhouse experiment was designed as previously described[17] with some modifications to allow randomization of theplant-growing containers on two greenhouse benches and toadapt the units to a recirculating drip system rather than acontinuous flow system. Two individual recirculating unitswere used. Each unit consisted of six, 21-L tub, a 30-Lreservoir and a 50-L refill tank. Three black polyethylenepots (24.5 cm×10.2 cm×10.16 cm), with one plant in each,were placed in each of the six, 21-L tubs. Pots tapered

downward to allow free drainage of nutrient solution. Thenutrient solution was pumped out from the reservoir througha network drip irrigation system into each individual pot.Irrigation time and intervals were controlled by a timer toprovide for 5 min watering duration every 2 h between8:00 A.M. and 8:00 P.M. The nutrient solution contained thefollowing nutrients: 16 NH4NO3, 482.23 KNO3, 708.45 Ca(NO3)2∙4H2O, 52.82 (NH4)2HPO4, 81.6 KH2PO4, 1.73H3BO3, 283.45 MgSO4∙7H2O, 0.049 CuSO4∙5H2O, 3.48K2SO4, 1.72 ZnSO4∙H2O, 0.61 (NH4)6Mo7O24∙4H2O, 0.79MnCl2, 28.93 Fe ethylenediaminetetraacetic acid (EDTA;98%) μg mL−1. Electrical conductivity was measuredweekly using a Horiba B-173 conductivity meter (SpectrumTechnologies, Inc., Plainfield, IL). Solution pH was adjustedmanually to 6.0±0.2 with H3PO4 or KOH every week.Nutrient solution temperatures were monitored every 30 min,using a temperature data logger (HOBO, ONSET ComputerCorporation).

Gerbera plant (G. jasemonii), cv. Timo, at the third-leafstage was transplanted into black polyethylene pot(24.5 cm×10.2 cm×10.16 cm), filled with perlite. Onemonth after transplant, the nutrient solution in each unitwas replaced with fresh nutrient solution. There were twogerbera experiments, and each consisted of two treatments:N-Serve®-amended and non-amended nutrient solution.N-Serve®, at 25 μg a.i. mL−1, was applied four times at3-week intervals. Each treatment included 18 plants.

Pepper

Experiments using pepper as the host were conducted in atemperature-controlled (24 to 34°C) greenhouse using fourseparate recirculating hydroponic units with top irrigation

Figure 1 Pepper (a) and ger-bera (b) plants in drip-irrigatedhydroponic units in a tempera-ture-controlled greenhouse

540 D. Pagliaccia et al.

system. Each hydroponic unit consisted of a plastic trough,located on a bench ca. 40 cm above ground level and connectedto a 50-L nutrient solution reservoir. Each trough contained sixpotted plants. The nutrient solution was pumped from thereservoir to the bench top and distributed to each plant viadrip tubing with 7.5 L/h emitters. Excess nutrient solutionfrom each pot drained by gravity back to the reservoir andwas recirculated. Plants were irrigated for 5 min every 2 hbetween 8:00 A.M. until 8:00 P.M. The solution pH wasadjusted manually to 6±0.2 with H3PO4 and KOH, everyweek. The electrical conductivity (EC 2±0.2 mS/cm2 forpepper) and the nutrient solution volume in reservoirs (25 L)were monitored twice a week and levels maintained asrequired. The final element concentrations in the nutrientsolution were as follows: Mg 48.8, S 64.5, P 62.1, K 241.5, N222.7, Ca 235.1, B 0.44, Cu 0.05, Cl 0.85, Mn 0.62, Mo 0.06,Zn 0.09, and Fe 2.5 μg ml−1.

Three-week-old pepper seedlings grown in Oasis Horti-cubes (Smithers-Oasis, Kent, OH) in a temperature-controlled incubator were transplanted to 11 cm×11 cm×12 cm plastic pots containing a commercial organic (peat-based) potting mix (Supersoil®, Rod McLellan Company,San Mateo, CA). Potted pepper plants were then placed introughs and grown for 4 days before treatment. There weretwo experiments each consisted of two treatments with sixplants per treatment: N-Serve®-amended and non-amendednutrient solution. N-Serve®, at 25 μg a.i. ml−1, was firstapplied 5 days after the start of the experiment andreapplied at days 12, 26, and 33. In all greenhouseexperiments, N-Serve® was added to the nutrient solution30 min before the start of the irrigation.

Monitoring Bacterial Populations in Nutrient Solutions

Twenty milliliter samples of the nutrient solution werecollected at different intervals (see Figs. 3 and 4) from thereservoir of each treatment. After serial tenfold dilutions,aliquots were plated in triplicates, using a spiral plater(Autoplate 4000-Spiral Biotech, Inc) onto 10-cm diameterpetri dishes containing King’s B medium (KB) forenumeration of the fluorescent pseudomonads and TrypticSoy Agar (TSA) for enumeration of the aerobic heterotro-phic bacteria. TSA was supplemented with 100 μg ml−1

cycloheximide (Calbiochen), and KB agar was supple-mented with 50 μg ml−1 cycloheximide and 75 μg ml−1

penicillin (Calbiochen). Colonies were counted after 24 hincubation at 28°C. Fluorescent pseudomonads werecounted under UV light.

Bacterial Collection and Identification

Eighty-five colonies of predominant bacteria growing onTSA medium were collected from the N-Serve®-amended

(50 isolates) and non-amended (35 isolates) nutrientsolutions from the gerbera and pepper experiments. Petridishes containing 50–100 single colonies were chosen, andcolonies were randomly selected from a different sector ofeach plate. The isolated colonies were purified by repeatedstreaking onto King B agar. Purified isolates were stored atroom temperature on KB agar plates or maintained forlong-term storage at −70°C in nutrient broth (DifcoLaboratories) supplemented with 15% glycerol. Eighty-fivebacteria isolates were identified using the BIOLOG™ GNplates and the Biolog Microlog2 4.20 software database(Biolog, Inc, Hayward, CA, USA). Furthermore, themetabolic patterns of the carbon source utilization profilesobtained with BIOLOG GN plates (Biolog Inc., Hayward,CA) were compared to understand if the enhanced P. putidapopulation after amendment is represented by a single ormultiple strains.

Genomic DNA Extraction

Samples of the gerbera nutrient solution from the reservoirs(1 L of hydroponic nutrient solution) from each treatmentwere collected onto sterile nitrocellulose membranes of twodifferent sizes, 0.22 and 8 μm (Millipore, nitrocellulose)using a sterile vacuum aspirator filtering apparatus. Thefilter-concentrated bacterial cells were resuspended in 6 mlbuffer 50 mM Tris–HCl pH 9, 20 mM EDTA, 400 mMNaCl, 0.75 M sucrose and stored until extraction procedurein −80°C freezer. Genomic DNA was extracted frombacterial suspension and pure cultures using the DNeasyTissue Kit (Qiagen, Hilden, Germany). The purity and theconcentration of the DNA template was verified byelectrophoresis in a 1% agarose gel and stained in ethidiumbromide solution.

PCR Amplification and T-RFLP Analysis

Bacterial 16S rRNA genes were amplified by polymerasechain reaction (PCR) with the primers FAM-63F, 5′-CAGGCCTAACA CATGCAAGTC-3′ labeled at the 5′end with 6-carboxyfluorescein dye, and 1389R, 5′-ACGGGCGGTGTG TACAAG-3′ [39] after the thermalprofile described by Bertin et al. [3]. Fluorescently labeledPCR products were purified with QIAquick PCR purifica-tion kit (Qiagen), and 100 ng were digested with 10 U ofthe restriction enzymes RsaI or HhaI (Invitrogen, Italia) forat least 4 h at 37°C. The digested products (2 μl) weremixed with 19.5 μl of deionized formamide and 0.5 μl ofROX-labeled GS500 internal size standard (Applied Bio-systems) and denaturated for 5 min at 95°C before capillaryelectrophoresis on ABI Prism 310 Genetic Analyzer(Applied Biosystems). Electropherogram analysis wasperformed using GeneScan Analysis 3.1 software using


the local southern size calling method. The computationalmethod developed by Abdo et al. [1] was used to determinea baseline, for indentification of the “true” peaks inelectropherograms and to compare electropherogram bin-ning fragments of similar size.

To investigate the distribution of bacterial diversity inthe analyzed samples, the peak relative intensity wasstatistically elaborated. Pairwise similarity between wholecommunities were analyzed by calculating Jaccard coeffi-cient Sj ¼ W= a1 þ a2 �Wð Þ� �

and Whittaker index ofassociation Sw ¼ 1�Σ=bi1 � bi2==2ð Þ, where W is thenumber of peaks shared between population 1 and 2; a1and a2 are the total number of different peaks in popula-tion 1 and 2; bi1 and bi2 are the relative fluorescence of eachith peak in samples 1 and 2 [30]. The Shannon–Weaver index H ¼ �ΣPi logPið Þ and Simpson index1� D ¼ 1�ΣP2

i

� �were determined according the de-

scription given in [30], where Pi is the relative fluorescenceof each ith peak. The values of ecological indices obtainedfor amended and non-amended samples were compared byusing Student’s t test.

DNA Sequencing and Phylogenetic Analysis

Genes, 16S rRNA, were amplified as described before fromgenomic DNA samples obtained from pure cultures of thestrains isolated from the hydroponic nutrient solution.Unlabeled PCR products, purified as described above, werecloned using the pGEM-T Easy Vector System (Promega)according to the manufacturer’s instructions and plasmidDNA, purified using a Wizard Plus SV Mini Prep kit(Promega), was subjected to cycle sequencing using theM13 primers and the BigDye Terminator Cycle SequencingReady Reaction kit (Applied Biosystems). The DNA sequen-ces were bidirectionally resolved on an ABI Prism 310 in asequencing mode. Nucleotide sequences were assembled andcompared with the sequences in the RDP database (Ribo-somal Database Project) to identify the closest relatives.

The 16S rRNA gene sequences obtained in this study areavailable from the EMBL nucleotide sequences databaseunder Accession number AM259176, AM259177,AM259178.

In Vitro Substrates Utilization

Bacterial Strains Used for In Vitro Studies

The following bacterial strains were isolated from thehydroponic nutrient solution: isolate 32, identified usingBIOLOG™ system as Aeromonas hydrophila, isolate 34,identified as Aeromonas caviae, isolate 36 and isolate Gerridentified as Pseudomonas putida by nucleotide sequencingof 16S ribosomal DNA.

Inocula for all in vitro experiments were prepared bygrowing the isolates first in LB broth (EMD ChemicalsInc.) at 28°C in a shaker (120 rpm) for 24 h, then in KBagar for 48 h. Finally, loops were transferred into 10 ml of0.01 M MgSO4 solution. These cells were washed twice inMgSO4 and CFU ml−1 assessed at 660 nm using Genesys10vs Spectrophotometer (Spectronic Unicam). These sus-pensions were then diluted to desired population densitiesfor in vitro studies.

Effect of N-Serve® on the Indigenous Bacterial Populations

Four, 0.5-L samples of gerbera nutrient solution werecollected from a nontreated reservoir in the greenhouseand then amended in the laboratory with 0, 25, 50 or100 μg a.i. ml−1 N-Serve®. After 30 min, aliquots wereplated in triplicates on KB agar and TSA agar thenincubated at 28°C for 24 h, and bacterial counts recordedas described above.

Effect of the Inert and Active Ingredients of N-Serve®on the Indigenous Fluorescent Pseudomonads

The active ingredient (i.e., nitrapyrin) and inert ingredients(i.e., xylene and 1,2,4-trimethylbenzene) of N-Serve®,either alone or in combination, were used as amendmentsto assess their effect on the bacterial population in thenutrient solution.

The nutrient solution with no chemical amendmentsserved as the control. Nitrapyrin (90%, N-ServeTG®) at12.5 μg a.i. ml−1 and/or xylene (99.7%, Mallinkrodt Baker,Inc.) at 20 μg ml−1 mixed with 1,2,4-trimethylbenzene(98%, Sigma-Aldrich, Inc) at 20 μg ml−1 were added to0.5 L of non-sterile nutrient solution collected from thehydroponic pepper recirculating system and incubated at28°C with continuous agitation on a rotary shaker(100 rpm). Fluorescent pseudomonad population in thechemically amended nutrient solutions were monitoreddaily for 6 days on King’s B agar as described above.The concentrations of the inert ingredients used in theabove experiments approximated those in the commercialformulation of N-Serve®.

Effect of N-Serve® on Pseudomonas Putida Gerr in SterileNutrient Solution and Sterile Distilled Water

To study the effect of N-Serve® on the growth of P.putida Gerr (a member of P. putida population selectivelyenhanced in gerbera experiment) in the absence ofcompetition from other bacteria, the bacterium was addedto sterile nutrient solution and sterile distilled water.Nutrient solution was collected from a nontreated reservoirof a hydroponic pepper experiment and filter sterilized by


passing through 8 and 0.22 μm filters (Millipore, nitrocel-lulose). N-Serve® (25 μg a.i. mL−1) was added to 0.5 L ofboth the sterile water and sterile nutrient solution andseeded with 4.4 log CFU ml−1 of P. putida. The treatmentswere incubated at 28°C with continuous agitation on arotary shaker (100 rpm). Bacterial populations weremeasured daily by plating serial dilutions onto King’s Bagar with a spiral plater (Autoplate 4000—Spiral Biotech,Inc.). The plates were incubated at 28°C and, after 48 h,colonies were counted under UV light.

Growth of Three Indigenous Bacterial Strains, IntroducedAlone or in Mixture, in Sterile Nutrient Solution Amendedwith N-Serve®

Three dominant and indigenous isolates of bacteria recov-ered from the nutrient solution [(i.e., A. hydrophila (isolate32), A. caviae (isolate 34) and an unidentified fluorescentpseudomonad (isolate 36)] were individually assessed fortheir ability to grow in sterile nutrient solution amendedwith and without N-Serve® at 25 μg a.i. ml−1. The nutrientsolution for these experiments was collected from a hydro-ponic pepper experiment and filter sterilized as describedabove. Inocula were prepared as previously described andthen added individually, in triplicates, to 200 ml of sterilepepper solution to give a final concentration of ca. 3.1 (logCFU ml−1). After 24 h, bacterial populations were enumer-ated as described previously.

To investigate the behavior of the isolates in a mixture, A.hydrophila, A. caviae, and the fluorescent pseudomonad,were each added at ca. 3.7 log CFU ml−1 to 200 ml of sterilepepper solution amended with N-Serve® (25 μg a.i. ml−1).The same concentration of this mixture, added to sterilenutrient solution served as a control. After 24 h, bacterialpopulations in the various treatments were enumerated asdescribed previously.

Nitrapyrin Experiment with Cu–CAS Solution and Fe–CASAgar Assays

To determine whether nitrapyrin a.i. is a metal chelator-siderophore and, more specifically, if it is a copper and ironchelator, we used both the Cu–CAS solution essay and theFe–CAS agar essay.

The Cu–CAS assay is based on ligand exchange, i.e., Cuis exchanged from the chrome azurol S (CAS) complex tothe siderophore. The experiment was performed according toReichman and Parker [48]. For the essay, 2.5 ml of samplesolution, prepared by adding N-Serve® at 12.5 μg a.i. ml−1

to deionized water, was mixed in triplicates to 2.5 ml of theCu–CAS reagent. Deionized water (2.5 ml) was used asreference blanks. The tubes were shaken briefly by handand the absorbance (Abs) determined at 582 nm. The

Cu–CAS essay from Reichman and Parker is a well-validated test that allows to detect ligand (i.e. siderophore)concentration in the sample solution between 30 nM and200 nM. The 25 μg ml−1concentration of nitrapyrin used inour greenhouse experiment corresponded to 270.7 nM; there-fore, after consultation with the authors, we chose the moreappropriate concentration of 12.5 μg ml−1 (=135.3 nM) forthe CAS test.

The Fe–CAS agar assay is based on the same principledescribed above except that the chelating activity ismeasured by the orange halo that develop in the agararound the metal chelator-siderophore as the Fe isremoved from the Fe–CAS dye complex, which givesthe medium its characteristic blue color. The methodologyused to prepare the Fe–CAS agar assay came fromAlexander and Zuberer [2]. For the essay, 10 μl of thepure product N-Serve® and 10 μl of the pure activeingredient nitrapyrin were spotted on the CAS agar and48 h later color change was examined.

Bacterial Siderophore Detection with Fe–CAS Solutionand Fe–CAS Agar Essays

A representative strain of P. putida no. 40 that wasselectively enhanced in the gerbera and pepper experiment,was assessed for its ability to produce siderophores in low-iron MM9 liquid medium [2] and in MM9 liquid mediumenriched with either 17.9 μM of FeEDTA (Sigma Chem-icals) or Fe-8hydroxyquinoline (Fe-8OHQ) and amendedwith or without N-Serve® at 25 μg a.i. ml−1. FeEDTAconcentration used was close to that which is present in ourhydroponic nutrient solution. Fe-8OHQ was used togetherwith FeEDTA to evaluate siderophore production with twodifferent metal chelators. To minimize the presence of non-chelated 8OHQ, equimolar amount of FeCl36H2O (Sigma)was added to 8OHQ-supplemented medium (99%, Cole-Parmer Instrument Company) [15]. The concentration ofsiderophore in supernatant was determined using a Fe–CASassay solution similar to that first described by Schwyn andNeilands [51] and modified by Alexander and Zuberer [1].The casamino acid (Fisher Scientific) stock solution usedfor the MM9 medium was deferred [51] by extraction for36 h with 3% (w/w) 8-hydroxyquinoline in chloroformsolution. After separation of the two phases, the aqueoussolution was washed with fresh chloroform until no furthercolor could be transferred to the organic phase. For theexperiment, the isolate was grown in KB agar medium for48 h at 30°C. The cells were harvested, and the pellets werewashed and adjusted to 0.25 OD, as described earlier andthen added individually to 50 ml Falcon tubes with 25 ml ofsterile MM9 medium [2] to give a final concentration of ca.5 log CFU ml−1. There were six treatments with treereplicates per treatment: (1) MM9 with 25 μg a.i. ml−1


N-Serve® with no Fe; (2) MM9 with no N-Serve® and noFe; (3) MM9 with 25 μg a.i. ml−1 N-Serve® with 17.9 μMof FeEDTA; (4) MM9 with no N-Serve® with 17.9 μM ofFeEDTA; (5) MM9 with 25 μg a.i. ml−1 N-Serve® with17.9 μM of Fe-8OHQ; (6) MM9 with no N-Serve® with17.9 μM of Fe-8OHQ. The treatments were incubated at29°C with continuous agitation on a rotary shaker(180 rpm). During incubation, cell density was monitoredat 660 nm. When stationary phase was reached (40 to 60 h),samples were taken, the cell were spun down at 10,000 rpmfor 10 min, and the supernatant was filter-sterilized(0.22 μm, Millipore). The concentration of siderophore inthe supernatant was measured by mixing 0.5 ml of modifiedCAS assay solution with 0.5 ml supernatant with 10 μl ofshuttle solution. The absorbance was measured at 630 nmafter 10 min [2].

Furthermore, ten additional strains of P. putida weretested for siderophores production on Fe-CAS agar. Fe-CAS agar, described above, was prepared following themethodology by Alexander and Zuberer [2]. CAS plateswere spot inoculated with bacterial strains and observed forthe development of orange halo against dark blue back-ground around the colonies after 48 h of incubation at 30 C.

In Vitro Effect of P. putida Culture Supernatanton P. capsici and P. aphanidermatum Mycelium Growth

The culture supernatant of P. putida strain #40, obtainedfrom the siderophore detection experiment described above,were screened in vitro for their antagonistic activity towardsmycelial growth of P. capsici and P. aphanidermatum. Sixmilliliters supernatant of each treatment was poured intriplicates in a 6-cm Petri dish to which two fungal discs(3 mm) from pre-grown (2–3 days) culture of P. capsici andP. aphanidermatum were incorporated separately and incu-

bated at room temperature. In addition, fresh MM9 mediumsupernatant with no culture, amended or not with 17.9 μM ofFeCl36H2O, served as controls. After 4 and 2 days, radialcolony growth of the fungus was measured using a Zeisscompound microscope (×25). Mycelium growth was thenrated on a scale from 0 (no growth) to 10 (luxurious growth).The experiment was repeated at least twice.

Data Analyses

All data were log10 transformed before statistical analysis.Data on the bactericidal effect of N-Serve® on indigenouspseudomonads and AHB populations (Fig. 6) were analyzedby the one-way repeated measures analysis of variance,followed by the all Pairwise Multiple Comparison Proce-dures (Holm-Sidak method). In vitro studies on the growthstimulatory effects of the active versus inert ingredients ofN-Serve® (Fig. 7) on indigenous bacteria, 72 h aftertreatment, were analyzed by one-way analysis of variancefollowed by the all Pairwise Multiple Comparison Proce-dures (Holm-Sidak method). Data on the individual bacterialability to grow on N-Serve® and the data on the competitionassays (Tables 3 and 4) were analyzed by three-way analysisof variance followed by the Student–Newman–Keuls pair-wise multiple comparison procedures. Data from thebacterial siderophore detection with Fe-CAS solution (after40 and 60 h of incubation) from treatments 1 (MM9 with25 μg a.i. ml−1 N-Serve® with no Fe) and 2 (MM9 with noN-Serve® and no Fe) were analyzed by t test (at P=0.05).

The observed effects were considered statistically sig-nificant when the calculated P value is below 0.001. Alldata were analyzed using SigmaStat 3.0 statistical softwarepackage (SPSS Science, Chicago, IL). All experimentswere repeated at least once, and results from a singlerepresentative experiment were chosen for presentation.

Figure 2 Fluorescent pseudo-monad population in therecycled nutrient solution attime zero (103 CFU ml−1; (a)and 72 h (106 CFU ml−1) afterN-Serve® treatment (b)


Results

Greenhouse Studies

Monitoring Bacterial Populations in the Nutrient Solutions

Before the addition of N-Serve®, the recirculating nutrientsolution used to irrigate gerbera plants contained ca. 3.6 logCFU ml−1 of aerobic heterotrophic bacteria (AHB; Fig. 3a)and an undetectable level of fluorescent pseudomonads(Fig. 3b). However, within 72 h after the addition ofN-Serve®, the average population of fluorescent pseudo-monads in amended treatments (ca. 6 log CFU ml−1) weretwo to three orders of magnitude greater than that in non-amended treatments (ca. 3 log CFU ml−1; Figs. 2 and 3b).The population surge occurred after each consecutiveaddition of N-Serve® to the nutrient solution. The averagefluorescent pseudomonad population accounted for ca. 80%(ranging from 65 to 100%) of the total AHB population inthe N-Serve®-amended solutions but only 2% (range from0.3 to 4.3%) in the non-amended nutrient solution. Similarsignificant increases in the resident fluorescent pseudomo-nad population, compared to the non-amended controltreatment, were also consistently recorded in the pepperexperiments after the addition of N-Serve® to the recircu-lating nutrient solution (Fig. 4).

Bacterial Identification

Among the 50 isolates of dominant bacteria, randomlyselected (on TSA medium) from N-Serve®-amendednutrient solutions, 94% were fluorescent pseudomonads.Forty-five isolates (90%) were identified using Biolog™,

with similarity indices of 75% (ranging from 70–91%), asP. putida biotype A and two (4%) as Pseudomonas putidabiotype B. Three nonfluorescent isolates (6%) wereidentified with Biolog™ as Stenotrophomonas maltophilia.Moreover, 16S rDNA sequences of eight selected isolateswere analyzed and identified as P. putida (four isolates), P.fluorescens (two isolates), and S. maltophila (two isolates).Comparative analysis of these sequences with the RDPdatabase showed significant similarities with 16S rRNAgene sequences of plant commensal Pseudomonas, bacte-rial strains with antifungal, antioomycete, and algicidaeactivities, polycyclic aromatic hydrocarbons (PAHs), andpesticides degrading bacteria, environmental and clinical

Figure 3 Mean population densities of aerobic heterotrophic bacteria(a) and fluorescent pseudomonads (b) in a greenhouse experimentwith gerbera as the host plant. Treatments consisted of nutrientsolution amended and not amended (control) with N-Serve® every

3 weeks. Arrows indicate the timing of N-Serve® applications (25 μga.i. ml−1). Error bars represent the SE of the mean (n=3) wherevariation was great enough to be presented

Figure 4 Mean population densities of fluorescent pseudomonads ina greenhouse experiment with pepper as the host plant. Treatmentsconsisted of nutrient solution amended and not amended (control)with N-Serve®. Arrows indicate the timing of N-Serve® applications(25 μg a.i. ml−1). Error bars represent the SE of the mean (n=3)where variation was great enough to be presented


isolates of Pseudomonas; Sab values (similarity coefficient forquery and matching sequences) were between 0.85 than 0.99.Furthermore, the comparison between the metabolic profilesof the 45 P. putida biotype A isolates suggested thatN-Serve® amendment selectively enhance a single strain.Specifically, all isolates had a very similar metabolic profilewith little difference between the specific carbon sourcesutilized and differed mainly in the succinamic acid utilization.

In contrast, of the 35 randomly selected isolates (onTSA medium) of dominant bacteria, recovered from thenon-amended nutrient solution, none were identified as P.putida. Fifty-seven percent (20 isolates) could not beidentified to species using Biolog™, as the similarityindices were below 50%. The remaining 43% (15 isolates)

of dominant isolates were identified, with similarity indicesof 67% (ranging from 53 to 91%), as Pseudomonassynxantha (five isolates), Pseudomonas fulva (three iso-lates), Pseudomonas fluorescens biotype C (one isolate), A.caviae DNA group 4 (two isolates), A. hydrophila DNAgroup 1 (two isolates), Delfia acidovorans (one isolate),and Comamonas testosteroni (one isolate).

Community Structure Analysis

The distribution and the structure of the bacterial communityin response to the addition to N-Serve® to the recirculatingnutrient solution were evaluated using a molecular approach.The analysis of terminal restriction fragment length polymor-

Figure 5 T-RFLP profiles from RsaI (a) and HhaI (b) analysis of 16SrDNA gene PCR products amplified from DNA isolated fromamended and not amended gerbera nutrient solution, 3 days after N-

Serve® (25 μg a.i. ml−1) addition. Comparison of the relativeabundance of T-RFs after RsaI and HhaI digestion between notamended (white) and amended samples (black) (c)


phism (T-RFLP) of 16S rRNA genes has proven to provide aculture-independent method to compare microbial communi-ties and presumptively identify abundant members. Fluores-cently labeled universal primers that anneal to conservedregion of 16S rRNA genes in Bacteria (see experimentalprocedures) gave amplification products of the expected sizein all the samples analyzed. Data from T-RFs patternsgenerated with RsaI or HhaI as restriction enzymes werecombined to achieve a more accurate characterization of themicrobial communities in the nutrient solutions. As illustratedin Fig. 5, differences in profiles, and changes in numbers ofdiscernible peaks, could be seen among the samples fromsolution amended or not with N-Serve®. To identify “true”peaks and compare electropherograms binning fragments ofsimilar size, the method developed by Abdo et al. [1] for theanalysis of T-RFLP data was used. The results obtained wereshown in Fig. 3c where the data were normalized bycalculating the relative peak areas. RsaI and HhaI digestionof fragments amplified with universal primers for Bacteriagenerated 12 peaks. The predominant RsaI T-RFs in controlsamples from untreated nutrient solution were sized 383 bpand in profiles obtained with HhaI digestion, the major peakhad a size of 79 bp. The T-RFs pattern of samples amendedwith N-Serve® was less complex than the one generated withnon-amended samples and showed the presence of a peakthat alone account for more than 59% of the total peak area ofthe sample (608 bp RsaI profiles, 170 bp HhaI profiles).

The whole profile analysis of T-RFLP data showed that thefragments that were represented in T-RFs profiles of samplesamended with N-Serve® were also present in non-amendedsamples, but the relative abundance of each phylotype wassignificantly different. The Jaccard coefficient, which indi-cates the similarities between T-RFs patterns considering thepresence/absence of the bands, had a value of 0.42 (RsaI) and0.58 (HhaI); the Whittaker coefficient, which takes intoconsideration the relative intensity of each band, had lowervalues (0.34 and 0.42, respectively), supporting the hypoth-esis that the structure of the bacterial communities wasaffected by the addition of N-Serve®.

To describe the changes in the dominance among thephylotypes, the ecological diversity indices of Shannon–Weaver and Simpson were calculated (Table 1). Diversity

indices for both enzymes were not significantly different(P>0.13), indicating that the diversity of the samples wasindependent from the restriction enzyme used to generatethe T-RFs profiles. On the contrary, indices in samplesamended with N-Serve® were significantly lower thanthose obtained in non-amended samples (P<0.01). Thesedata suggest that the addition of N-Serve® decreases theprokaryotic diversity in the sample, thus reducing thespecies richness and evenness and enhancing a singlephylotype of the bacterial consortium.

T-RFLP analysis, performed on pure culture DNA ofGerr isolate generated a single T-RF of 170 bp after HhaIdigestion and 608 bp fragment after RsaI digestion (datanot shown), indicating that the dominant fluorescent speciesin the nutrient solution samples amended with N-Serve®matched with the P. putida isolate.

In Vitro Substrates Utilization

Effect of N-Serve® on the Indigenous Bacterial Populations

The population densities of resident AHB decreasedsignificantly (50–85%; Fig. 6a), compared to those in thenon-amended nutrient solution, within 30 min after theaddition of 25, 50 and 100 μg a.i. ml−1of N-Serve® tothe nonsterile nutrient solution. In contrast, no significantreduction in population density of indigenous fluorescentpseudomonads (IFP; Fig. 6b) was observed in the nonsterilenutrient solution amended with 25 and 50 μg a.i. ml−1 ofN-Serve®. However, a significant decrease (40%) in theIFP population was observed in the nutrient solutionamended with 100 μg a.i. ml−1of N-Serve®.

Effect of the Inert and Active Ingredients of N-Serve®on the Indigenous Fluorescent Pseudomonads

Seventy-two hours after the addition of the inert ingredients(xylene and 1,2,4-trimethylbenzene) or a mixture of the activeingredient (nitrapyrin) plus the inert ingredients (xylene and1,2,4-trimethylbenzene) to the non-sterile nutrient solution inthe laboratory, the mean population densities of indigenousfluorescent pseudomonads increased from 2.7 to 4.6 log CFUml−1 and from 2.70 to 6.13 log CFU ml−1, respectively(Fig. 7). Significant increases in population densities in themixture-treated solution was observed compared to the inertingredients amendment. In contrast, the population densityof indigenous fluorescent pseudomonads decreased duringthe same time frame in both the unamended and nitrapyrin-amended treatments. The drop in population densities wasnot considerably different (although with a significant greatersurvival with nitrapyrin after 72 h) but both were signif-icantly different from the nitrapyrin plus inert ingredient andinert ingredient treatments.

Table 1 Values for ecological diversity indices of Shannon-Weaver(H) and Simpson (1-D) obtained by using the data of the HhaI andRsaI generated T-RFLP

HhaI T-RFs pattern RsaI T-RFs pattern

H 1-D H 1-D

Not amended 2.36±0.21 0.90±0.07 2.15±0.13 0.85±0.09Amended 1.36±0.08 0.62±0.05 0.91±0.10 0.44±0.03


Effect of N-Serve® on Pseudomonas Putida Gerr in SterileNutrient Solution and Sterile Distilled Water

Within 72 h after the addition of N-Serve® to sterile nutrientsolution seeded with P. putida Gerr (see experimentalprocedures), bacterial counts increased from 4.40 to 5.98log CFU ml−1 but decreased rapidly in the absence of

N-Serve® (Table 2). In contrast, bacterial counts decreasedfrom 4.4 log CFU ml−1 to non-detectable levels within 24 hafter the addition of N-Serve® to sterile distilled water. Adecrease in the population of P. putida Gerr also occurred inthe non-amended sterile water, but the decrease was less thanin the N-Serve®-amended treatment.

Growth of Three Indigenous Bacterial Isolates, IntroducedAlone or in Mixture, in Sterile Nutrient Solution Amendedwith N-Serve®

Significant increases in the populations of each of three(i.e., two nonfluorescent and one fluorescent) members of

Figure 6 Effect of N-Serve® at 25, 50 and 100 μg a.i. ml−1 onpopulation densities of indigenous aerobic heterotrophic bacteria (a)and fluorescent pseudomonads (b) after 30 min. Nutrient solutionsamples from greenhouse gerbera experiment were treated in vitro.AHB populations were determined by plating a dilution series of eachtreatment onto Tryptic soy agar and fluorescent pseudomonads by

plating onto King’s B agar followed by incubation at 25°C for 2 days.Values are the means of three replicates. Bars with different letters aresignificantly different at P=0.01 according to the all pairwise multiplecomparison procedures (Holm-Sidak method). Error bars representthe SE of the mean (n=3) where variation was great enough to bepresented

Figure 7 Mean population densities of indigenous fluorescentpseudomonads in nutrient solution collected from hydroponic pepperrecirculating system. Nutrient solution samples were treated in vitroand compared with controls. Treatments consisted of control nutrientsolutions receiving no chemical, and/or amended with nitrapyrin(12.5 μg a.i. ml−1) and/or xylene and 1,2,4-trimethylbenzene (20 +20 μg ml−1) at 0 time. At 3 days time, points with different letters aresignificantly different at P=0.01 according to the all pairwise multiplecomparison procedures (Holm-Sidak method). Error bars representthe SE of the mean (n=3) where variation was great enough to bepresented

Table 2 Mean populations of fluorescent pseudomonads Gerr in steriledistilled water (SDW) and sterile gerbera hydroponic nutrient solution(SNS) amended or not amended with N-Serve® at 25 μg a.i. ml-1

Treatments a Bacterial concentration (Log CFU mL−1) b

Incubation period (h)

0 0.5 24 48 72 96

SDW + P. putidaGerr c

4.40 4.32 3.98 2.79 2.08 –

SDW + P. putidaGerr + N-Serve®

4.40 3.19 0 0 0 –

SNS + P. putida Gerr 4.40 – – 1.30 1.30 1.30SNS + P. putidaGerr + N-Serve ®

4.40 – – 5.06 5.98 4.90

a N-Serve® (25 μg a.i. ml−1 ) was applied to 500 ml of sterile distilledwater (SDW) or sterile nutrient solution (SNS)bMean of three replicates samplescPseudomonas putida Gerr strain was added at ±4.40 log CFU ml−1


the dominant indigenous bacterial isolates (which wereisolated from unamended nutrient solution) occurred insterile nutrient solution amended with N-Serve® (Table 3).No increase occurred in the absence of N-Serve®. Inaddition, a significant increase in the population, relative tothe control, was also observed after the addition of all threebacteria as a mixture to sterile nutrient solution amendedwith N-Serve®. However, the fluorescent pseudomonadaccounted for almost 100% of the observed populationincrease (Table 4). Comparison analysis confirmed thatthere was no significant difference (P=0.561) between totalbacteria population and fluorescent Pseudomonas isolate 36population, 24 h after N-Serve® amendment.

Nitrapyrin Experiment with Cu–CAS Solution and Fe–CASAgar Assays

The results from the Cu–CAS solution essay demonstratedthat nitrapyrin a.i. (used as N-Serve® commercial productand pure nitrapyrin a.i.) is unable to remove Cu from itsCAS complex, as no change in the Cu–CAS solution colorand in the absorbance value (Abs582<0.01) were detected(data not shown). The results from the Fe–CAS agar essayalso established that nitrapyrin was unable to remove Fefrom its CAS complex, as no change in color of the CAS-blue agar occurred.

Bacterial Siderophore Detection with Fe–CAS Solutionand Fe–CAS Agar Essays

P. putida growth and accumulation of siderophores in MM9medium with or without Fe, and or N-Serve®, are shown inFig. 8a and b, respectively. P. putida grew well in MM9supplemented with either 17.9 μM of FeEDTA or Fe-8OHQ. No difference in population growth was observedwith and without N-Serve® amendment. However, inmedium without added Fe, P. putida growth was consider-ably lower (about one third compared to the Fe addedMM9; Fig. 8a).

The accumulation of siderophores in the medium wasmonitored by measuring the absorbance of culture filtratesat 630 nm (Fig. 8b). No siderophores accumulated incultures grown in MM9 supplemented with either 17.9 μMof FeEDTA or Fe-8OHQ, and no difference in siderophoreaccumulation was observed with and without N-Serve®amendment. A considerable amount of siderophores wereaccumulated in MM9 containing no added Fe (Fig. 8b).After 40 h incubation, there was a significant difference(P=0.049; with a mean value of 67 vs 49 U) in thesiderophores excreted by P. putida in MM9 media contain-ing no added Fe but amended with N-Serve® compared to

Table 3 Bacterial population densities of pure cultures of strains innutrient solution amended with N-Serve®

Treatmentsa Bacterialconcentration(Log CFU ml−1)b


0 h 24 h

N-Serve® NDd NDAeromonas hydrophila isolate 32 c 3.12 ae 3.09 aAeromonas caviae isolate 34 3.23 a 3.18 aPseudomonas spp. isolate 36 3.17 a 3.09 aN-Serve® + Aeromonas hydrophila isolate 32 3.10 a 5.36 cN-Serve® + Aeromonas caviae isolate 34 3.22 a 5.01 bN-Serve® + Pseudomonas spp. isolate 36 3.19 a 5.32 c

a N-Serve® (12.5 μg a.i. ml−1 ) was applied to 200 ml of sterilehydroponic nutrient solutionbMean of three replicates samplesc Strains were previously isolated from hydroponic nutrient solutionfrom a pepper greenhouse experimentd No bacteria were detected in sterile hydroponic nutrient solutionamended with N-Serve®e Value followed by the same letter are not significantly different atP=0.05

Table 4 Population densities of a bacterial mixture in nutrient solution amended with N-Serve®

Treatmentsa Total population (Log CFU ml−1)b Fluorescent Pseudomonas spp. population (Log CFU ml−1)


0 h 24 h 0 h 24 h

N-Serve® NDc ND ND NDBacterial mixtured 4.43 b 4.40 b 3.74 a 3.70 aN-Serve® + Bacterial mixture 4.47 be 5.93 c 3.76 a 5.88 c

a N-Serve® (12 μg a.i. ml−1 ) was applied to 200 ml of sterile hydroponic nutrient solutionbMean of three replicate samplesc No bacteria were detected in sterile hydroponic nutrient solution amended with N-Serve®d Bacterial mixture was made of A. hydrophila isolate 32, A. caviae isolate 34 and fluorescent Pseudomonas spp. isolate 36. Each strain was addedat ±1.3 log CFU ml−1 concentratione Value followed by the same letter are not significantly different at P=0.05


the unamended N-Serve® medium, but no difference(P=0.81) was observed at 60 h incubation.

Siderophores production could also be detected on Fe–CAS agar from all ten additional strain of P. putida tested,where a strong orange halo (0.5–1 cm) was produced by allstrains after 48 h incubation.

In Vitro Effect of P. putida Culture Supernatant on P.capsici and P. aphanidermatum Mycelium Growth

The results represented in Fig. 9 showed that the superna-tant from P. putida strain no. 40 in all treatments had asignificant antagonistic activity towards P. capsici and P.aphanidermatum mycelium growth indicating that antifun-gal factors other than siderophores were produced by P.putida. Culture supernatant from P. putida grown witheither 17.9 μM of FeEDTA or Fe-8OHQ had a slightly

higher suppression of hyphal growth than that with noadded Fe (Fig. 9a and b). The difference was morepronounced in the experiment with P. capsici (Fig. 9b). Incontrast, supernatant not supplemented with Fe showed agreater inhibitory activity toward sporangium production(data not showed).

Discussion

Fluorescent pseudomonads have historically been associatedwith suppression of root diseases caused by numerous fungaland fungal-like pathogens. However, inconsistency inperformance has been a problem. This has been attributed,in part, to the inability of the biocontrol bacteria to maintaincritical threshold populations necessary for sustained bio-control activity. It is becoming increasingly evident thatbiocontrol bacteria will not be effective for sustained periodof time unless the environment is modified to make it moreconducive for their growth and survival [7].

Our results indicate that a nitrogen stabilizer (N-Serve®,Dow Agrosciences) selectively and significantly enhanced

Figure 8 Growth (a) and accumulation of siderophore (b), by batchcultures of P. putida strain no. 40. Siderophore concentration ispresented as percent values, calculated by using the followingformula: Ar � Asð Þ=Ar½ �100, where Ar is the recorded value of thereference solution (r) and As is the recorded value of the samplesolution at 660 nm absorbance (Abs). Each point is the mean of threereplicates

Figure 9 Antifungal activities of the cell-free culture supernatant ofP. putida strain no. 40 (Pp) (grown on MM9 medium for 3 days)against mycelial cultures of P. aphanidermatum (a) and P. capsici (b).Mycelial growth is represented as values between 0 (no growth) to 10(luxurious growth). Each value is the mean of two experiments, each,with three replicates±SD


the fluorescent pseudomonad population in the recirculatingnutrient solution reservoirs in both hydroponically growngerbera and pepper plants (Figs. 3 and 4). Our resultssupport and extend the concept of selective enhancement ofa specific microorganism in a root/soil microbial commu-nity by providing an exotic substrate. The use of nutritionalamendments has previously been reported to enhance theefficacy of several bacterial and fungal biological controlagents of fungal diseases [21]. For example, Colbert et al.[6–8] demonstrated that amending soil with salicylateselectively increased the metabolic activity and populationsof an introduced strain (PpG7) of P. putida. Yamada [69]demonstrated that amending soil with methionine enhancedthe population of P. putida AP-1 which increased suppres-sion of Fusarium wilt of tomato. Similarly, olive oil, whenadded to the recirculating nutrient solution of hydroponi-cally grown peppers, stimulated the metabolic activity (i.e.,rhamnolipid production) and population density of anintroduced strain (R4) of P. aeruginosa which resulted inthe control of P. capsici [58]. Utilization of exoticsubstrates to enhance specific microorganisms is notrestricted to root/soil microcosms but has also been reportedfor foliar microcosms [11, 23, 44, 67–69].

In contrast to the above studies in which the microor-ganism intended for enhancement were intentionally intro-duced into the desired microcosm, the bacterium enhancedin our studies existed naturally, at low population densities,in the nutrient solution. The enhanced bacterium wassubsequently identified, using BIOLOG™ identificationtest and comparison of 16S rRNA gene sequences (Fig. 5),as a P. putida strain. This bacterium, either directly orindirectly, was associated with disease suppression in ourhydroponic cultural systems [40]. A role in biologicalcontrol by P. putida has been shown [18, 65].

Coupled with the enhancement of P. putida population, areduction in species richness (i.e., a lower bacterialdiversity expressed as number of T-RFs), occurred in theN-Serve®-amended compared to the unamended nutrientsolution (Fig. 5a and b). It has been reported that diversityof microbial communities can be affected by organicpollutants and that organic chemicals reduce microbialdiversity in different environments [14, 47]. Our in vitrostudy supports these conclusions. Specifically, we observedthat the bacterial population bursts in our study waspreceded (within 30 min after amending the nutrientsolution with N-Serve®) by early growth suppression ofthe indigenous AHB population, while no effect wasobserved on the fluorescent pseudomonad population(Fig. 6). After this lag period, the fluorescent pseudomonadpopulation was selectively enhanced, as evidenced by datafrom the greenhouse (Figs. 3 and 4). These results areconsistent with a previous report from Liu and Suflita [33],which observed that the composition of the indigenous

microbial population in the soil and ground water will adaptto the presence of pollutants mainly because bacteria thatare able to use these compounds substrates as a nutrientsource would be able to proliferate and became dominant.In support of this hypothesis, our competition assays(Table 4) provided evidence that indigenous N-Serve®-tolerant bacteria out-compete others for available resourcesand the non-pseudomonad bacteria were the most impaired.

Thus, given that N-Serve® contains inert ingredientssuch as xylene and 1,2,4-trimethylbenzene, we hypothe-sized that the increase in fluorescent pseudomonadpopulation could be the result of the degradation ofthose inert ingredients and their utilization as a carbonsource. Our in vitro studies on the growth stimulatoryeffects of the active versus inert ingredients of N-Serve®(Fig. 5) on indigenous bacteria in the nutrient solutionconfirmed that the inert components of N-Serve®, used ascarbon source, were responsible of a significant percentageof the bacterial population increase, and more specifically,the fluorescent pseudomonad population. Preliminaryresults (Dr. M. Ruzzi, Personal communication: data notshown) further established that the P. putida, selectivelyenhanced by N-Serve®, could effectively metabolize xyleneand 1,2,4-trimethylbenzene. These results are consistentwith previous reports that pseudomonads are metabolicallyversatile and are capable of utilizing many natural andxenobiotic compounds, including xylene and benzene [14,29, 35, 36, 55, 60].

Unclear and more ambiguous is the role of nitrapyrin onthe bacterial growth stimulatory effect. Our laboratorystudies showed that nitrapyrin, in contrast to the inertingredients, cannot be used by the bacteria as a sole carbonsource, however prevented the rapid decline of thepseudomonad populations over time (Fig. 7). Furthermore,when the active and inert ingredients were mixed together,the increase of the bacterial population was significantlyhigher than those recorded in response to the inertingredients alone (Fig. 7). The reduced competition due tothe bactericidal activity of nitrapyrin (Fig. 6) and the use ofthe inert ingredients from pseudomonads as carbon source(Fig. 7, and preliminary results (Dr. M. Ruzzi, Personalcommunication) seemed to play a major role in the bacterialgrowth stimulation. On the other hand, results did notsupport the possibility that metal chelator, by providingpseudomonads with essential metals, give the bacteriumgrowth advantage over other bacterial species, as in ourexperiments (Cu–CAS solution and Fe–CAS agar assays),nitrapyrin failed to exhibit chelating activity for copper andiron. However, it is possible that nitrapyrin chelates copperand iron, but perhaps with lower binding affinity than thatof CAS.

The selectively enhanced pseudomonads growth patternsdescribed above occurred in experimental system using


recycled hydroponic nutrient solution, which containedinorganic nutrients, organic, and inorganic matter. When apure culture of fluorescent P. putida isolate Gerr grown indistilled water was treated with N-Serve®, a rapid decreasein bacterial growth was observed. When the same strainwas grown in a sterile hydroponic nutrient solution treatedwith N-Serve®, the pseudomona population densitiesincreased by three log units (Table 2). It has been reportedthat pseudomonas have diverse metabolic pathways con-trolled by a variety of plasmid and chromosomal geneticmechanisms often affected by environmental factors [27,43, 57]. More specifically Zhou and Crawford [70] reportedthat microbial degradation of organic compounds isstrongly influenced by physical and chemical factors suchas nutrients, temperature, oxygen, salinity, pressure, wateractivity, pH, and chemical composition. The importance ofthe inorganic nutrients (i.e. N and P) on the bacterialstimulation for pollutant degradation has also been reported[13, 49, 53]. In our study, the nutrient-rich hydroponicsolution, in combination with the stable pH, E.C., andoxygen supply most likely stimulated and sustained theexpression of the catabolic mechanisms of the amendedaromatic compounds resulting to the observed P. putida-enhanced populations. The results of the competitive assaywith A. hydrophila and A. caviae against fluorescentpseudomonad further supported the metabolic durabilityand versatility of the florescent pseudomonads in theN-Serve® amended hydroponic environment, as the non-pseudomonas isolates were able to increase their populationonly in the absence of the florescent pseudomonads(Table 4).

Regardless of the mechanism of selective populationincrease of fluorescent pseudomonads observed in ourstudies, this type of increase is typically seen to stimulatebiological control activity [18, 41, 42]. Numerous studieshave shown that siderophores from the fluorescent pseudo-monads, like pyoverdine and other antifungal metaboliteshave been implicated as contributing to biocontrol activityagainst soilborne diseases. [31, 34, 41, 46].

Results presented in this paper showed that the P. putidastrains, whose population tends to be enhanced byN-Serve®, were strong producer of pioverdines on KBagar and siderophores in iron depleted media (Fig. 8b).However, in iron rich media, the in vitro studies demon-strated that P. putida culture supernatant, free of side-rophores, also had a significant inhibitory activity against P.capsici and P. aphanidermatum (Fig. 9a and b) suggestingthat this bacteria may also be producing other antifungalmetabolites (i.e., antibiotics or other compounds). Theresults of our greenhouse experiments [40] indicated thatthe selectively enhanced fluorescent pseudomonads weregood root colonizers as well, and that one P. putida straincontrolled plant disease in hydroponic systems. Further-

more, antagonistic effect towards P. capsici and P.aphanidermatum was also obtained with pure cultures ofP. putida on our nutrient solution agar assays [37] (data notshown), indicating that the strain secretes indeed antifungalfactors. These findings suggest that this bacterium couldhave a broad-spectrum antagonistic activity. However,further investigations are needed to understand the extent towhich parameters such as iron availability and carbon sourcecan be controlled in hydroponic systems to take advantage ofthe biocontrol mechanism and thus makes it effective. Weneed also to establish the extent to which each or all of themechanism contribute to the efficacy of biological control inour system. These control mechanisms could be optimized ina controlled system such as the hydroponics in which abioticfactors (i.e., carbon source, mineral availability, etc.) can bemonitored and manipulated. However, determining the typeand levels of metabolites that are being produced in situ aredifficult, compared to in vitro condition [41]. Therefore, whilewe clearly established a role for the iron in our in vitroexperiment, we could not establish clearly that the samephenomenon is occurring in situ; i.e., the biocontrolmediated by either siderophores, antibiotics, or inducedresistance. In several studies, iron has been implicated as astimulant of production of a variety of antifungal metaboliteswhen added as an amendment [12, 22, 56]. However, inother studies, pseudomonads were found to secrete side-rophores in iron-depleted system [41].

The significance of these studies could be of importantvalue. First, it could help to solve one of the main problemsassociated with inconsistency in performance of a biocon-trol agent, namely, maintenance of a high populationdensity of the selected biocontrol agent. The residentmicroflora in many, if not most, habitats could alreadycontain prospective members that could function asantagonists to other microorganisms in the same habitat ifselectively enhanced. This hypothesis is consistent withprinciples established for bioremediation by Kuiper et al.[24, 25], in which naturally occurring bacteria were selectedfor the ability to both degrade a pollutant and colonize plantroots. In our case, this principle could be used to insure theselection of a fluorescent pseudomonad population from therhizosphere and the nutrient solution, for beneficial purposesuch as biocontrol, biofertilization, and phytostimulation.

Second, N-Serve® was identified as a specific chemicalwhich, when applied at low concentrations, can selectivelyenhance specific microbes among a milieu of others. Moreefficient compounds with similar properties could beidentified or refined for their selective enhancement forspecific biological purposes. While hydroponic systemswith recirculating nutrient solutions offer optimal condi-tions for bacterial growth, i.e., the relatively high concen-trations of root exudates, non-limiting minerals, highaeration and incubation temperatures of 2030°C, the same


principle may also have application in field agriculture. Forexample, these growth-stimulating chemicals could beapplied via the drip irrigation system and used tomanipulate specific indigenous microbial populations forplant health management.

Acknowledgements We wish to thank J. Adaskaveg, D.H. Ferrin, I.J. Misaghi, P. Barghini, and G. Vidalakis for editorial assistance andfor technical help; H. R. Azad, for support with the BIOLOG test; D.Parker, A. Seyfferth and D. E. Crowley for their fruitful discussionand collaboration. This research was partially supported by grantsfrom the USDA/ARS Floral and Nursery Crop Research Initiative andby a grant from ISPESL, Italy (Project B96-2DIPIA/03).

References

1. Abdo Z, Schuette UME, Bent SJ, Williams CJ, Forney LJ, JoyceP (2006) Statistical methods for characterizing diversity ofmicrobial communities by analysis of terminal restriction frag-ment length polymorphisms of 16S rRNA genes. Environ Micro-biol 8:929–938

2. Alexander DB, Zuberer DA (1991) Use of chrome azurol Sreagents to evaluate siderophore production by rhizospherebacteria. Biol Fertil Soils 12:39–45

3. Bertin L, Colao MC, Ruzzi M, Marchetti L, Fava F (2006)Performances and microbial features of an aerobic packed-bedbiofilm reactor developed to post-treat an olive mill effluent froman anaerobic GAC reactor. Microb Cell Fact 5:16

4. Bloemberg GV, Lugtenberg BJJ (2001) Molecular basis of plantgrowth promotion and biocontrol by rhizobacteria. Curr OpinPlant Biol 4:343–350

5. Campbell NER, Aleem MIH (1965) The effect of 2-chloro,6-(trichloromethyl) pyridine on the chemoautotrophic metabolism ofnitrifying bacteria. Antonie van Leeuwenhoek 31:124–136

6. Colbert SF, Isakeit T, Ferri M, Weinhold AR, Hendson M, SchrothMN (1993a) Use of an exotic carbon source to selectively increasemetabolic activity and growth of Pseudomonas putida in soil.Appl Environ Microbiol 59:2056–2063

7. Colbert SF, Isakeit T, Ferri M, Weinhold AR, Hendson M, SchrothMN (1993b) Enhanced growth and activity of a biocontrolbacterium genetically engineered to utilize salicylate. ApplEnviron Microbiol 59:2071–2076

8. Colbert SF, Schroth MN, Weinhold AR, Hendson M (1993c)Enhancement of population densities of Pseudomonas putidaPpG7 in agricultural ecosystems by selective feeding with thecarbon source salicylate. Appl Environ Microbiol 59:2064–2070

9. Cornelis P, Matthijs S (2002) Diversity of siderophore-mediatediron uptake systems in fluorescent pseudomonads: Not onlypyoverdines. Environ Microbiol 4:787–798

10. Cortese MS, Paszczynski A, Lewis TA, Sebat JL, Borek V,Crawford RL (2002) Metal chelating properties of pyridine-2,6-bis(thiocarboxylic acid) produced by Pseudomonas ssp. and thebiological activities of the formed complexes. Biometals 15:103–120

11. Davis RF, Backman PA, Rodriguez-Kabana R, Kokalis-Burelle N(1992) Biological control of apple fruit diseases by Chaetomiumglobosum formulations containing cellulose. Biol Control 2:118–123

12. Dwivedi D, Johri BN (2003) Antifungals from fluorescentpseudomonads: Biosynthesis and regulation. Curr Sci 85:1693–1703

13. Evans FF, Rosado AS, Sebastian GV, Casella R, Machado PLOAet al (2004) Impact of oil contamination and biostimulation on thediversity of indigenous bacterial communities in soil microcosms.FEMS Microbiol Ecol 49:295–305

14. Fahy A, Lethbridge G, Earle R, Ball AS, Timmis KN, McGenityTJ (2005) Effects of long-term benzene pollution on bacterialdiversity and community structure in groundwater. EnvironMicrobiol 7:1192–1199

15. Geels FP, Schmidt EDL, Schippers B (1985) The use of 8-hydroxy-quinoline for the isolation and prequalification of plantgrowth-stimulating rhizosphere pseudomonads. Biol Fertil Soils1:167–173

16. Goring CAI (1962) Control of nitrification by 2-chloro,6-(trichloromethyl) pyridine. Soil Sci 93:211–218

17. Grusak MG, Pezeshgi S (1994) Uniformly 15N-labeled soybeanseeds produced for use in human and animal nutrition studies:Description of a recirculating hydroponic growth system andwhole plant nutrient and environmental requirements. J Sci FoodAgric 64:223–230

18. Haas D, Defago G (2005) Biological control of soil-bornpathogens by fluorescent pseudomonads. Nature Rev Microbiol3:307–319

19. Hooper AB, Terry KR (1973) Specific inhibitors of ammoniaoxidation in Nitrosomonas. J Bacteriol 115:480–485

20. Iizumi T, Misumoto M, Nakamura K (1998) A bioluminiscenceassay using Nitrosomonas europaea for rapid and sensitivedetection of nitrification inhibitors. Appl Environ Microbiol64:3656–3662

21. Ji P, Wilson M (2003) Enhancement of population size of abiological control agent and efficacy in control of bacterial speckof tomato through salicylate and ammonium sulfate amendments.Appl Environ Microbiol 69:1290–1294

22. Keel C, Voisard C, Berling CH, Kahr G, Défago G (1989) Ironsufficiency, a prerequisite for the suppression of tobacco blackroot rot by Pseudomonas fluorescens strain CHA0 undergnotobiotic conditions. Phytopathology 79:584–589

23. Kokalis-Burelle N, Backman PA, Rodriguez-Kabana R, PloperLD (1992) Potential for biological control of early leafspot ofpeanut using Bacillus cereus and chitin as foliar amendments. BiolControl 2:321–328

24. Kuiper I, Bloemberg GV, Lugtenberg BJJ (2001a) Selection of aplant-bacterium pair as a novel tool for rhizostimulation ofpolycyclic aromatic hydrocarbon-degrading bacteria. Mol PlantMicrobe Interact 10:1197–1205

25. Kuiper I, Kravchenko LV, Bloemberg GV, Lugtenberg BJJ (2001b)Pseudomonas putida strain PCL1444, selected for efficient rootcolonization and naphthalene degradation, effectively utilizes rootexudates components. Mol Plant Microbe Interact 7:734–741

26. Kuiper I, Lagendijk E, Bloemberg G, Lugtenberg BJJ (2004)Rhizoremediation: A beneficial plant-microbe interaction. MolPlant-Microbe Interact 17:6–15

27. Kunz DA, Chapman PJ (1981) Isolation and characterization ofspontaneously occurring TOL plasmid mutants of Pseudomonasputida HS1. J Bacteriol 146:952–964

28. Laskowski DA, O’Melia FC, Griffith JD, Regoli AJ, YoungsonCR, Goring CAI (1975) Effect of 2-chloro-6- (trichloromethyl)pyridine and its hydrolysis product 6-chloropicolinic acid on soilmicroorganisms. J Environ Qual 4:412–417

29. Leahy JG, Tracy KD, Eley MH (2003) Degradation of mixtures ofaromatic and chloroaliphatic hydrocarbons by aromatic hydrocar-bon-degrading bacteria. FEMS Microbiol Ecol 43:271–276

30. Legendre P, Legendre L (1998) Numerical Ecology. Elsevier,Amsterdam

31. Lemanceau P, Bakker PAHM, De Kogel WJ, Alabouvette C,Schippers B (1992) Effect of pseudobactin 358 production byPseudomonas putida WC358 on suppression of Fusarium wilt of


carnations by nonpathogenic Fusarium oxysporum Fo47. ApplEnviron Microbiol 58:2978–2982

32. Lewis TA, Paszczynski A, Wylie SWG, Jeedigunta S, Lee CH,Crawford RL (2001) Carbon tetrachloride dechlorination by thebacterial transition metal chelator pyridine-2,6-bis(thiocarboxylicacid). Environ Sc Technol 35:552–559

33. Liu S, Suflita JM (1993) Ecology and evolution of microbialpopulations for bioremediation. Trends Biotechnol 11:344–352

34. Loper J, Buyer JS (1991) Siderophore in microbial interactions onplant surfaces. Mol Plant Microbe Interact 4:5–13

35. Mallakin A, Ward OP (1996) Degradation of BTEX compoundsin liquid media and in peat biofilters. J Ind Microbiol 16:309–318

36. Molina L, Ramos C, Duque E, Ronchel MC, Garcia JM, Wyke L,Ramos JL (2000) Survival of Pseudomonas putida KT2440 in soiland in the rhizosphere of plants under greenhouse and environ-mental conditions. Soil Biol Biochem 32:315–321

37. Ongena M, Daayf F, Jacques P, Thonart P, Benhamou N, PaulitzTC, Cornelis P, Koedam N, Belanger RR (1999) Protection ofcucumber against Pythium root rot by fluorescent pseudomonads:Predominant role of induced resistance over siderophores andantibiosis. Plant Pathol 48:66–76

38. Ongena M, Giger A, Jacques P, Dommes J, Thonart P (2002)Study of bacterial determinants involved in the induction ofsystemic resistance in bean by Pseudomonas putida BTP1. Eur JPlant Pathol 108:187–196

39. Osborn AM, Moore ERB, Timmis KN (2000) An evaluation ofterminal-restriction fragment length polymorphism (T-RFLP)analysis for the study of microbial community structure anddynamics. Environ Microbiol 2:39–50

40. Pagliaccia D, Ferrin D, Stanghellini ME (2007) Chemo-biologicalsuppression of root-infecting zoosporic pathogens in recirculatinghydroponic systems. Plant Soil 299:163–179

41. Pal, KK, McSpadden Gardener, B (2006) Biological control ofplant pathogens. The Plant Health Instructor DOI 10.1094/PHI-A-2006–1117–02

42. Paulitz TC, Belanger RR (2001) Biological control in greenhousesystems. Ann Rev Phytopathol 39:103–133

43. Phoenix P, Keane A, Patel A, Bergeron H, Ghoshal S, Lau PCK(2003) Characterization of a new solvent-responsive gene locus inPseudomonas putida F1 and its functionalization as a versatilebiosensor. Environ Microbiol 5:1309–1327

44. Ploper LD, Backman PA, Rodriguez-Kabana R (1992) Enhancednatural biological control of apple fruit diseases by applications ofbiopolymers. Biol Control Tests 7:3

45. Powell SJ, Prosser JI (1986) Effect of copper on inhibition bynitrapyrin of growth of Nitrosomonas europaea. Curr Microbiol14:177–179

46. Raaijmakers JM, Leeman M, van Oorschot MMP, Van der Sluis I,Schippers B, Bakker PAHM (1995) Dose-response relationshipsin biological control of Fusarium wilt of radish by Pseudomonasspp. Phytopathology 85:1075–1080

47. Ramos JL, Diaz E, Dowling D, de Lorenzo V, Molin S, O’Gara Fet al (1994) The behavior of bacteria designed for biodegradation.Nat Biotechnol 12:1349–1356

48. Reichman SM, Parker DR (2007) Critical evaluation of threeindirect assays for quantifying phytosiderophores released by theroots of Poaceae. Eur J Soil Sci 58:844–853

49. Röling WFM, Milner MG, Jones DM, Lee K, Daniel F, SwannellRJP, Head IM (2002) Robust hydrocarbon degradation anddynamics of bacterial communities during nutrient-enhanced oilspill bioremediation. Appl Environ Microbiol 68:5537–5548

50. Runia, W (1996) Disinfection of recirculation water from closedproduction systems. In: van Os EA (Ed.) Proceedings of theSeminar on Closed Production Systems. pp 20–24

51. Schwyn B, Neilands JB (1987) Universal chemical assay for thedetection and determination of siderophores. Anal Biochem 160:47–56

52. Sebat JL, Paszczynski AJ, Cortese MS, Crawford RL (2001)Antimicrobial properties of pyridine-2,6-dithiocarboxylic acid, ametal chelator produced by Pseudomonas spp. Appl EnvironMicrobiol 67:3934–3942

53. Segura A, Duque E, Mosqueda G, Ramos JL, Junker F (1999)Multiple responses of Gram-negative bacteria to organic solvents.Environ Microbiol 1:191–198

54. Shattuck GE, Alexander M (1963) A differential inhibitor ofnitrifying microorganisms. Soil Sci Soc Am Proc 27:600–601

55. Shim H, Yang ST (1999) Biodegradation of benzene, toluene,ethylbenzene, and o-xylene by a coculture of Pseudomonas putidaand Pseudomonas fluorescens immobilized in a fibrous-bedbioreactor. J Biotechnol 67:99–112

56. Slininger PJ, Jackson MA (1992) Nutritional factors regulatinggrowth and accumulation of phenazine 1-carboxylic acid byPseudomonas fluorescens 2–79. Appl Microbiol Biotechnol37:388–392

57. Stallwood B, Shears J, Williams PA, Hughes KA (2005) Lowtemperature bioremediation of oil-contaminated soil using bio-stimulation and bioaugmentation with a Pseudomonas sp. frommaritime Antarctica. J Appl Microbiol 99:794–802

58. Stanghellini ME, Miller RM (1997) Biosurfactants: Their identityand potential efficacy in the biological control of zoosporic plantpathogens. Plant Dis 81:4–12

59. Stanghellini ME, Rasmussen SL (1994) Hydroponics: A solutionfor zoosporic pathogens. Plant Dis 78:1129–1138

60. Timmis KN (2002) Pseudomonas putida: A cosmopolitanopportunist par excellence. Environ Microbiol 4:779–781

61. Trenkel ME (1997) Improving Fertilizer Use Efficiency - Control-Release and Stabilized Fertilizers in Agriculture. InternationalFertilizer Industry Association, Paris

62. Van Loon LC, Bakker PAHM, Pieterse CMJ (1998) Systemicresistance induced by rhizosphere bacteria. Ann Rev Phytopathol36:453–483

63. Vannelli T, Hooper AB (1992) Oxidation of nitrapyrin to 6-chloropicolinic acid by the ammonia-oxidizing bacterium Nitro-somonas europaea. Appl Environ Microbiol 58:2321–2325

64. Van Os EA, Postma J (2000) Prevention of root diseases in closedsoilless growing systems by microbial optimization and slow sandfiltration. Acta Hortic 532:97–102

65. Walsh UF, Morrissey JP, O’Gara F (2001) Pseudomonas forbiocontrol of phytopathogens: From functional genomics tocommercial exploitation. Curr Opin Biotechnol 12:289–295

66. Weller DM, Raaijmakers JM, McSpadden Gardener BB, Thoma-show LS (2002) Microbial populations responsible for specificsuppressiveness to plant pathogens. Ann Rev Phytopathol40:309–348

67. Wilson M, Campbell HL, Ji P, Jones JB, Cuppels DA (2002)Biological control of bacterial speck of tomato under fieldconditions at several locations in North America. Phytopathology92:1284–1292

68. Wilson M, Savka MA, Hwang I, Farrand SK, Lindow SE (1995)Altered epiphytic colonization of mannityl opine-producingtransgenic tobacco plants by a mannityl opine-catabolizing strainof Pseudomonas syringae. Appl Environ Microbiol 61:2151–2158

69. Yamada M, Ogiso M (1997) Control of soil-borne diseases usingantagonistic microorganisms. IV. Study on the available substratesfor antagonistic bacterial strains to control Fusarium wilt oftomatoes. Res Bull Aichi-Ken Agric Res Cent 29:141–144

70. Zhou E, Crawford RL (1995) Effects of oxygen, nitrogen, andtemperature on gasoline biodegradation in soil. Biodegradation6:127–140


dx.doi.org/10.1094/PHI-A-2006-1117-02

dx.doi.org/10.1094/PHI-A-2006-1117-02

Selective Enhancement of the Fluorescent Pseudomonad Population After Amending the Recirculating...

Documents

Transcript of Selective Enhancement of the Fluorescent Pseudomonad Population After Amending the Recirculating...