RESUMO
The deferred antagonism technique has been utilized for several decades for detecting antibiosis activity. Most protocols require the elimination of antibiotic-producing cells by exposing them to chloroform vapour, UV radiation or filter sterilizing the filtrate steps that require additional time and expense to complete. We provide a modified approach to current soft agar overlay practices, which involves addition of antibiotics to the soft agar overlay to inhibit growth of the producer but not the indicator strain. This technique can be used to reproducibly and efficiently screen for antibiotic production with ease. We demonstrate the effectiveness of this technique with three bacterial systems: inhibition of the bacterial spot of tomato pathogen, Xanthomonas euvesicatoria, by its pathogenic competitor Xanthomonas perforans; and inhibition of the fire blight pathogen, Erwinia amylovora, by Pantoea vagans C9-1 or Pseudomonas fluorescens A506. SIGNIFICANCE AND IMPACT OF THE STUDY: Deferred antagonism assays are used commonly to observe antibiotic production by micro-organisms. Killing or removing the producer cells prior to introduction of the indicator strain is a standard practice but requires additional time and special handling procedures. We evaluated a modification of the assay, where the overlay medium is amended with an antibiotic to which the indicator strain is resistant and the producer strain is sensitive. This modification obviates extra steps to kill the producer strain prior to overlaying with the indicator strain and provides a rapid, consistent and cost-effective method to detect antibiosis.
Assuntos
Antibiose , Erwinia amylovora/fisiologia , Técnicas Microbiológicas/métodos , Pantoea/fisiologia , Pseudomonas fluorescens/fisiologia , Xanthomonas/fisiologia , Solanum lycopersicum/microbiologia , Doenças das Plantas/microbiologia , Xanthomonas/crescimento & desenvolvimentoRESUMO
Antibiotics are essential for control of bacterial diseases of plants, especially fire blight of pear and apple and bacterial spot of peach. Streptomycin is used in several countries; the use of oxytetracycline, oxolinic acid and gentamicin is limited to only a few countries. Springtime antibiotic sprays suppress pathogen growth on flowers and leaf surfaces before infection; after infection, antibiotics are ineffective. Antibiotics are applied when disease risk is high, and consequently the majority of orchards are not treated annually. In 2009 in the United States, 16,465 kg (active ingredient) was applied to orchards, which is 0.12% of the total antibiotics used in animal agriculture. Antibiotics are active on plants for less than a week, and significant residues have not been found on harvested fruit. Antibiotics have been indispensable for crop protection in the United States for more than 50 years without reports of adverse effects on human health or persistent impacts on the environment.
Assuntos
Antibacterianos/farmacologia , Bactérias/efeitos dos fármacos , Produtos Agrícolas/microbiologia , Doenças das Plantas/prevenção & controle , Erwinia amylovora/efeitos dos fármacos , Erwinia amylovora/patogenicidade , Doenças das Plantas/microbiologiaRESUMO
Pantoea agglomerans E325, the active ingredient in a commercial product for fire blight control, was previously shown in vitro to produce a unique alkaline- and phosphate-sensitive antibiotic specific to Erwinia amylovora. Antibiosis was evaluated as a mode of antagonism on flower stigmas using two antibiosis-deficient mutants. On King's medium B, mutants E325ad1 and E325ad2 have stable smooth-butyrous or hypermucoid colony morphologies, respectively, and the parental strain E325 exhibits phenotypic plasticity with predominantly hypermucoid colonies accompanied by slower-growing, smooth-butyrous colonies. Mutants were tested against E. amylovora on stigmas of detached flowers of crab apple (Malus mandshurica) in growth chambers and apple (Malus domestica) in the orchard. Epiphytic fitness of the antibiosis-negative mutants was similar or greater than the parental strain as determined by relative area under the population curve (RAUPC). In laboratory and orchard trials, both mutants had significantly lower inhibitory activity against the pathogen (i.e., less reduction of E. amylovora RAUPC) compared with the parental strain. E325 and the mutants caused similar decreases in pH in a broth medium, indicating that acidification, which was previously reported as a possible mechanism of pathogen inhibition on stigmas, is not directly related to antibiosis. In this study we provide the first evidence for E325 antibiosis involved in E. amylovora growth suppression on apple flower stigmas.
Assuntos
Erwinia amylovora/crescimento & desenvolvimento , Flores/microbiologia , Malus/microbiologia , Pantoea/fisiologia , Doenças das Plantas/microbiologia , Aldose-Cetose Isomerases/genética , Aldose-Cetose Isomerases/metabolismo , Antibiose , Proteínas de Bactérias/genética , Proteínas de Bactérias/metabolismo , Mutagênese Insercional , Oxirredutases , Pantoea/enzimologia , Pantoea/genética , Análise de Sequência de DNARESUMO
Mixtures of biological control agents can be superior to individual agents in suppressing plant disease, providing enhanced efficacy and reliability from field to field relative to single biocontrol strains. Nonetheless, the efficacy of combinations of Pseudomonas fluorescens A506, a commercial biological control agent for fire blight of pear, and Pantoea vagans strain C9-1 or Pantoea agglomerans strain Eh252 rarely exceeds that of individual strains. A506 suppresses growth of the pathogen on floral colonization and infection sites through preemptive exclusion. C9-1 and Eh252 produce peptide antibiotics that contribute to disease control. In culture, A506 produces an extracellular protease that degrades the peptide antibiotics of C9-1 and Eh252. We hypothesized that strain A506 diminishes the biological control activity of C9-1 and Eh252, thereby reducing the efficacy of biocontrol mixtures. This hypothesis was tested in five replicated field trials comparing biological control of fire blight using strain A506 and A506 aprX::Tn5, an extracellular protease-deficient mutant, as individuals and combined with C9-1 or Eh252. On average, mixtures containing A506 aprX::Tn5 were superior to those containing the wild-type strain, confirming that the extracellular protease of A506 diminished the biological control activity of C9-1 and Eh252 in situ. Mixtures of A506 aprX::Tn5 and C9-1 or Eh252 were superior to oxytetracycline or single biocontrol strains in suppressing fire blight of pear. These experiments demonstrate that certain biological control agents are mechanistically incompatible, in that one strain interferes with the mechanism by which a second strain suppresses plant disease. Mixtures composed of mechanistically compatible strains of biological control agents can suppress disease more effectively than individual biological control agents.
Assuntos
Erwinia amylovora/fisiologia , Pantoea/fisiologia , Doenças das Plantas/microbiologia , Doenças das Plantas/prevenção & controle , Pseudomonas fluorescens/fisiologia , Pyrus/microbiologia , Interações Hospedeiro-PatógenoRESUMO
The biological control agents Pseudomonas fluorescens A506 and Pantoea vagans C9-1 were evaluated individually and in combination for the suppression of fire blight of pear or apple in 10 field trials inoculated with the pathogen Erwinia amylovora. The formulation of pathogen inoculum applied to blossoms influenced establishment of the pathogen and the efficacy of biological control. Pantoea vagans C9-1 suppressed fire blight in all five trials in which the pathogen was applied as lyophilized cells but in none of the trials in which the pathogen was applied as freshly harvested cells. In contrast, Pseudomonas fluorescens A506 reduced disease significantly in only one trial. A mixture of the two strains also suppressed fire blight, but the magnitude of disease suppression over all field trials (averaging 32%) was less than that attained by C9-1 alone (42%). The two biological control agents did not antagonize one another on blossom surfaces, and application of the mixture of A506 and C9-1 to blossoms resulted in a greater proportion of flowers having detectable populations of at least one bacterial antagonist than the application of individual strains. Therefore, the mixture of A506 and C9-1 provided less disease control than expected based upon the epiphytic population sizes of the antagonists on blossom surfaces. We speculate that the biocontrol mixture was less effective than anticipated due to incompatibility between the mechanisms by which A506 and C9-1 suppress disease.
Assuntos
Pantoea/patogenicidade , Doenças das Plantas/microbiologia , Pseudomonas fluorescens/patogenicidade , Malus/microbiologia , Pantoea/crescimento & desenvolvimento , Doenças das Plantas/prevenção & controle , Pseudomonas fluorescens/crescimento & desenvolvimento , Pseudomonas fluorescens/fisiologia , Pyrus/microbiologia , Árvores/microbiologiaRESUMO
As a prerequisite to infection of flowers, Erwinia amylovora grows epiphytically on stigmas, which provide a conducive habitat for bacterial growth. Stigmas also support growth of several other bacterial genera, which allows for biological control of fire blight; although, in practice, it is very difficult to exclude E. amylovora completely from this habitat. We investigated the dynamics of growth suppression of E. amylovora by comparing the ability of virulent and avirulent strains of E. amylovora to compete with each other on stigmas of pear, apple, and blackberry, and to compete with a co-inoculated mixture of effective bacterial antagonists. When strains were inoculated individually, virulent E. amylovora strain Ea153N attained the highest population size on stigmas, with population sizes that were approximately double those of an avirulent hrpL mutant of Ea153 or the bacterial antagonists. In competition experiments, growth of the avirulent derivative was suppressed by the antagonist mixture to a greater extent than the virulent strain. Unexpectedly, the virulent strain enhanced the population size of the antagonist mixture. Similarly, a small dose of virulent Ea153N added to inoculum of an avirulent hrpL mutant of Ea153 significantly increased the population size of the avirulent strain. A pathogenesis-gene reporter strain, Ea153 dspE::gfp, was applied to flowers and a subset of the population expressed the green fluorescent protein while growing epiphytically on stigmas of apple. These results are consistent with the hypothesis that virulent E. amylovora modifies the epiphytic habitat presented by the stigma through a pathogenesis-related process, which increases host resources available to itself and, coincidentally, to nonpathogenic competitors. Over nine orchard trials, avirulent Ea153 hrpL significantly suppressed the incidence of fire blight four times compared with six for the antagonist mixture. The degree of biological control achievable with an avirulent strain of E. amylovora likely is limited by its inability to utilize the stigmatic habitat to the same degree as a virulent strain.
Assuntos
Erwinia amylovora/patogenicidade , Flores/microbiologia , Controle Biológico de Vetores/métodos , Doenças das Plantas/microbiologia , Rosaceae/microbiologia , VirulênciaRESUMO
Pantoea agglomerans strain E325, a commercially available antagonist for fire blight of apple and pear, was originally selected through screening based on suppression of Erwinia amylovora on flower stigmas, but specific mechanisms of antagonism were unknown. Bacterial modification of pH was evaluated as a possible mechanism by analyzing stigma exudates extracted from 'Gala' apple stigmas. The pH values for field samples were only slightly lower than controls, but indicated a range (pH 5 to 6) conducive for antibiotic activity according to subsequent assays. Under low-phosphate and low-pH conditions, an antibacterial product of E325 with high specificity to E. amylovora was effective at low concentrations. A minimum of 20 to 40 ng of a ninhydrin-reactive compound purified using RP-HPLC caused visible inhibition in assays. Activity was heat stable and unaffected by amino acids, iron, or enzymes known to affect antibiotics of P. agglomerans. Antibiosis was diminished, however, under basic conditions, and with increasing phosphate concentrations at pH 6 and 7. Inhibition was not observed in media containing phosphate concentrations commonly used in antibiosis assays. We propose that E325 suppresses the fire blight pathogen not only by competing for nutrients on the stigma, but by producing an antibiotic specific to E. amylovora. Further work is necessary to substantiate that the compound is produced and active on flower stigmas.
Assuntos
Antibacterianos/biossíntese , Erwinia amylovora/genética , Frutas/microbiologia , Malus/microbiologia , Pantoea/efeitos dos fármacos , Antibacterianos/uso terapêutico , Erwinia amylovora/efeitos dos fármacos , Erwinia amylovora/patogenicidade , Flores/microbiologia , Concentração de Íons de Hidrogênio , Testes de Sensibilidade Microbiana , Pantoea/patogenicidade , Fosfatos/análise , Doenças das Plantas/microbiologiaRESUMO
We investigated adaptation of fire blight forecasting concepts to incorporate and optimize the use of biological agents for disease suppression. The effect of temperature on growth of the bacterial antagonists, Pseudomonas fluorescens A506 and Pantoea agglomerans C9-1S, and of the pathogen Erwinia amylovora153N, on pear and apple blossoms was evaluated in growth chamber and screenhouse experiments. New blossoms were inoculated with the strains and subsequent growth was measured over 96 h. Bacterial growth rates on blossoms were described as functions of temperature. A degree hour-based "bacterial growth index" (96-h moving total of degree hours >10°C) was created to assess conduciveness of orchard environments for antagonist colonization. A comparison of this index to a disease risk index indicated that biocon-trol treatments could be timed such that the antagonists could be expected to grow to an effective population size before the disease index shifted from "low" to "moderate" risk. For six pear- and apple-production areas of Oregon and Washington, regression of actual values of the bacterial growth and disease risk indices on index values derived from 4-day temperature forecasts resulted in coefficients of determination that averaged 0.75. The "bacterial growth index" and its estimation via temperature forecasts were incorporated into a decision matrix designed to guide optimal treatment timing.
RESUMO
ABSTRACT Fire blight, caused by Erwinia amylovora, is the most serious bacterial disease of pear and apple trees. Biological control with strains of Pantoea agglomerans (syn. Erwinia herbicola) may provide an effective disease management strategy for fire blight. Most strains of P. agglomerans evaluated for suppression of fire blight produce compounds that inhibit the growth of E. amylovora in culture. The role of these inhibitory compounds in fire blight suppression in orchard environments has not been studied. In seven field trials in Oregon, we compared the population dynamics and disease suppression with P. agglomerans Eh252, a strain that produces a single antibiotic, with its near-isogenic antibiotic-deficient derivative, strain 10:12. Water or suspensions of Eh252 or 10:12 (1 x 10(8) CFU/ml) were applied at 30 and 70% bloom to pear or apple trees. Aqueous suspensions of freeze-dried cells of E. amylovora (3 x 10(5) CFU/ml) were applied at full bloom. Additional trees were treated with streptomycin or oxytetracycline at 30 and 70% bloom and in some experiments, 1 day after application of the pathogen. Population sizes of Eh252 or 10:12 on pear blossoms were estimated by spreading dilutions of blossom washes on culture media. Average population sizes of Eh252 and 10:12 on blossoms ranged from 10(5) to 10(7) CFU, and in five of six trials, the relative area under the population curve of Eh252 was not significantly different than that of its derivative 10:12. Both Eh252 and 10:12 reduced the growth of the pathogen on blossoms compared with inoculated water-treated controls. Eh252 significantly decreased the incidence of fire blight in six of seven field trials compared with the incidence on water-treated trees, and 10:12 similarly reduced the incidence of fire blight in four of seven trials. In three of seven field trials, trees treated with Eh252 had a significantly lower incidence of fire blight compared with trees treated 3 with 10:12. Overall,3 Eh252 reduced the incidence of fire blight by 55 +/- 8%, 10:12 by 30 +/- 6%, streptomycin by 75 +/- 4%, and oxytetracycline by 16 +/- 14%. The effectiveness of strain 10:12 compared with water treatment indicates that other mechanisms (e.g., competitive exclusion or habitat modification) also contribute to disease suppression by P. agglomerans. The increased suppression of fire blight by the parental strain Eh252 compared with the antibiotic-deficient mutant 10:12 indicates that antibiosis is an important mechanism of biological control of fire blight.
RESUMO
Pseudomonas fluorescens Pf-5 is a soil bacterium that suppresses plant pathogens due in part to its production of the antibiotic pyoluteorin. Previous characterization of Pf-5 revealed three global regulators, including the stationary-phase sigma factor sigma(S) and the two-component regulators GacA and GacS, that influence both antibiotic production and stress response. In this report, we describe the serine protease Lon as a fourth global regulator influencing these phenotypes in Pf-5. lon mutants overproduced pyoluteorin, transcribed pyoluteorin biosynthesis genes at enhanced levels, and were more sensitive to UV exposure than Pf-5. The lon gene was preceded by sequences that resembled promoters recognized by the heat shock sigma factor sigma(32) (sigma(H)) of Escherichia coli, and Lon accumulation by Pf-5 increased after heat shock. Therefore, sigma(H) represents the third sigma factor (with sigma(S) and sigma(70)) implicated in the regulation of antibiotic production by P. fluorescens. Lon protein levels were similar in stationary-phase and exponentially growing cultures of Pf-5 and were not positively affected by the global regulator sigma(S) or GacS. The association of antibiotic production and stress response has practical implications for the success of disease suppression in the soil environment, where biological control organisms such as Pf-5 are likely to encounter environmental stresses.
Assuntos
Antibacterianos/biossíntese , Proteínas de Bactérias , Proteínas de Escherichia coli , Proteínas de Choque Térmico/metabolismo , Protease La , Pseudomonas fluorescens/enzimologia , Serina Endopeptidases/metabolismo , Fator sigma , Raios Ultravioleta , Proteases Dependentes de ATP , Sequência de Bases , Elementos de DNA Transponíveis , Proteínas de Ligação a DNA/genética , Proteínas de Ligação a DNA/metabolismo , Regulação Bacteriana da Expressão Gênica , Proteínas de Choque Térmico/genética , Resposta ao Choque Térmico , Dados de Sequência Molecular , Mutagênese Insercional , Fenóis , Fosfotransferases (Aceptor do Grupo Álcool)/metabolismo , Pseudomonas fluorescens/genética , Pseudomonas fluorescens/metabolismo , Pseudomonas fluorescens/efeitos da radiação , Pirróis , Tolerância a Radiação , Serina Endopeptidases/genética , Fatores de Transcrição/genética , Fatores de Transcrição/metabolismoRESUMO
ABSTRACT We evaluated effects of both physical and biological components of the environment on growth of Pantoea agglomerans on inoculated pear and apple blossoms and on spread of the bacterium to blossoms on non-inoculated trees. The center three rows of 0.35- to 0.5-ha blocks of four pear cultivars and four apple cultivars were sprayed with a suspension of streptomycin-resistant P. agglomerans strain C9-1S (C9-1S) at 20 to 60% and 60 to 90% bloom. Cultivars were chosen to create a sequence of continuous bloom from late March (d'Anjou pear) through mid-May (Red Rome apple). Each cultivar block was quartered into plots; two plots were treated twice with streptomycin sulfate near mid- and full bloom to suppress populations of indigenous bacterial epiphytes and the other two plots were treated with water. Colonization of blossoms by C9-1S and by indigenous bacterial epiphytes were monitored on inoculated trees and along transects of noninoculated trees. Immediately after spraying, C9-1S was detected principally on blossoms sampled from inoculated trees. As bloom progressed, trees up to 18 m from inoculated trees had high proportions of blossoms colonized by C9-1S. Streptomycin significantly (P = 0.05) reduced incidence of isolation and size of detectable populations of culturable bacteria (indigenous bacteria plus C9-1S) from pear blossoms in 1998 and from apple blossoms in both 1998 and 1999, but the antibiotic treatment did not affect incidence of isolation, size of detectable populations, or spread of C9-1S compared to the water-treated control in any experiment. Across all cultivars, relative area under the curve for size of detectable populations of C9-1S on inoculated trees and for incidence of isolation of C9-1S from noninoculated trees was positively correlated with mean degree hours per day during bloom (r= 0.61 to 0.73) and negatively correlated with the proportion of days with rain (r = -0.79 to -0.84). The results indicate that establishment and growth of C9-1S on pome fruit flowers was not strongly affected by streptomycin or by competition from indigenous bacterial epiphytes and, as with Erwinia amylovora, temperature is an important environmental variable affecting successful spread of this biological control agent from blossom to blossom.
RESUMO
ABSTRACT Pear blossoms were sampled during various stages of bloom in 1991 and 1992 from orchards at Cashmere, WA, and Corvallis and Medford, OR, for epiphytic populations of culturable bacteria. On stigmatic surfaces, bacteria were isolated from 2 to 32% of blossoms prior to petal expansion and from 47 to 94% of blossoms by petal fall. In general, a lower percentage of hypanthia than stigmas supported bacterial populations. Randomly selected bacteria isolated at population levels of >/=10(4) CFU/tissue were identified by fatty acid methyl ester analysis. Diverse genera of gram-negative and -positive bacteria were identified from the Medford and Cashmere field sites. Pseudomonas syringae and Pseudomonas viri-diflava were isolated from all sites and were the predominant species detected at Corvallis, where they were isolated from 28% of the blossoms sampled on a given date. Because most pear blossoms do not support detectable populations (>/=10(2) CFU/tissue) of culturable bacteria prior to petal expansion, we speculate that introduced biocontrol agents may become established with minimal competition from indigenous epiphytes at early bloom stages.
RESUMO
The influence of inoculum preparation on the establishment of bacterial antagonists that suppress fire blight and Erwinia amylovora on blossoms was evaluated. Aqueous suspensions of Pseudomonas fluorescens A506, E. herbicola C9-1R, or E. amylovora 153N were prepared from cells harvested from the surface of an agar medium or from cells that were lyophilized after culture under similar conditions. Bacterial suspensions (1 x 10(8) CFU/ml) were sprayed on pear and apple trees at 50% bloom near midday. The incidence of recovery (proportion of blossoms containing detectable populations) and the population sizes of the bacteria on individual blossoms with detectable populations were followed over a period of several days. Fluorescent microspheres (1 mum in diameter) were added to sprays at a concentration of 1 x 10(7) microspheres per ml to mark blossoms that were open during application of bacteria. After dilution-plating, the stigmas and styles of each blossom were examined for the presence of microspheres with an epifluorescence microscope. In three of five trials, bacteria applied as suspensions of lyophilized cells were recovered from a greater proportion of blossoms than bacterial cells harvested directly from culture media. Every blossom harvested within 6 days after spraying had microspheres present on the surfaces of the styles and stigmas; thus, lack of establishment of detectable populations, rather than escape of blossoms from spray inoculation, accounted for the differences in proportion of blossoms colonized by the different preparations of bacteria. The use of lyophilized cells in field trials decreased variability in the establishment of bacteria on blossoms.
RESUMO
Dispersal of the bacteria Pseudomonas fluorescens strain A506 and Erwinia herbicola strain C9-1S from treated to nontreated pear blossoms, and the effect of their spread on fire blight, were investigated in an orchard block of 10 rows containing 4 trees per row. Center rows of trees were sprayed with a mixture of P. fluorescens A506 and E. herbicola C9-1S at 30, 15, and 50% bloom in 1994, 1995, and 1996, respectively. Immediately after spraying, antagonists were detected only on treated blossoms. In 1994 and 1996, as bloom progressed, both P. fluorescens A506 and E. herbicola C9-1S were detected on nontreated blossoms located up to 4 rows (10 m) from the treated rows. In 1995, establishment of the antagonists on treated blossoms was poor and spread to nontreated trees was limited, apparently because of cold temperatures. Each year, honey bees were used to inoculate all trees with E. amylovora at 80% bloom. After full bloom in 1994 and 1996, the proportion of blossoms with E. amylovora populations >105 CFU per flower were highest in the outermost rows, and decreased linearly (P < 0.05) with proximity to treated rows. In 1994, diseased blossom clusters decreased significantly (P < 0.05) from the outermost rows to the treated rows, but there was no significant effect of distance on disease incidence in 1995 or 1996. Secondary colonization of blossoms by P. fluorescens A506 and E. herbicola C9-1S can play a role in disease suppression, but, among seasons, rates of secondary colonization by P. fluorescens A506 and E. herbicola C9-1S were variable, indicating that multiple applications of antagonists may be necessary to optimize biological control.
RESUMO
Suppression of the blossom-blight phase of fire blight is a key point in the management of this destructive and increasingly important disease of apple and pear. For blossom infection to occur, the causal bacterium, Erwinia amylovora, needs to increase its population size through an epiphytic phase that occurs on stigmatic surfaces. Knowledge of the ecology of the pathogen on stigmas has been key to the development of predictive models for infection and optimal timing of antibiotic sprays. Other bacterial epiphytes also colonize stigmas where they can interact with and suppress epiphytic growth of the pathogen. A commercially available bacterial antagonist of E. amylovora (BlightBan, Pseudomonas fluorescens A506) can be included in antibiotic spray programs. Integration of bacterial antagonists with chemical methods suppresses populations of the pathogen and concomitantly, fills the ecological niche provided by the stigma with a nonpathogenic, competing microorganism. Further integration of biologically based methods with conventional management of blossom blight may be achievable by increasing the diversity of applied antagonists, by refining predictive models to incorporate antagonist use, and by gaining an improved understanding of the interactions that occur among indigenous and applied bacterial epiphytes, antibiotics, and the physical environment.
RESUMO
Agrobacterium radiobacter K84 is an effective, commercially applied, biological control agent for the plant disease crown gall, yet little is known about the survival and dissemination of K84. To trace K84 in the environment, spontaneous antibiotic-resistant mutants were used. Growth rates and phenotypes of streptomycin- or rifampin-resistant K84 were similar to those of the parental K84, except the rifampin-resistant mutant produced less agrocin 84 as determined by bioassay. K84 and a strain of Agrobacterium tumefaciens established populations averaging 10(5) CFU/g in the rhizosphere of cherry and persisted on roots for 2 years. K84 established rhizosphere populations between 10(4) and 10(6) CFU/g on cherry, ryegrass, and 11 other herbaceous plants. Populations of K84 declined substantially in fallow soil or water over a 16-week period. K84 was detected in the rhizosphere of ryegrass located up to 40 cm from an inoculum source, indicating lateral dissemination of K84 in soil. In gall tissue on cherry, K84 established populations of 10(5) CFU/g, about 10- to 100-fold less than that of the pathogen. These data demonstrate that K84 persists for up to 2 years in a field environment as a rhizosphere inhabitant or in association with crown gall tissue.