Evaluation of bacteria isolated from rice for plant growth promotion ...

Download PDF
3 downloads 0 Views 899KB Size Report
Comment
Tika B. Adhikari, C.M. Joseph, Guoping Yang, Donald A. Phillips, and. Louise M. Nelson. Abstract: Of 102 rhizoplane and endophytic bacteria isolated from rice ...
916

Evaluation of bacteria isolated from rice for plant growth promotion and biological control of seedling disease of rice Tika B. Adhikari, C.M. Joseph, Guoping Yang, Donald A. Phillips, and Louise M. Nelson

Abstract: Of 102 rhizoplane and endophytic bacteria isolated from rice roots and stems in California, 37% significantly (P ≤ 0.05) inhibited the growth in vitro of two pathogens, Achlya klebsiana and Pythium spinosum, causing seedling disease of rice. Four endophytic strains were highly effective against seedling disease in growth pouch assays, and these were identified as Pseudomonas fluorescens (S3), Pseudomonas tolaasii (S20), Pseudomonas veronii (S21), and Sphingomonas trueperi (S12) by sequencing of amplified 16S rRNA genes. Strains S12, S20, and S21 contained the nitrogen fixation gene, nifD, but only S12 was able to reduce acetylene in pure culture. The four strains significantly enhanced plant growth in the absence of pathogens, as evidenced by increases in plant height and dry weight of inoculated rice seedlings relative to noninoculated rice. Three bacterial strains (S3, S20, and S21) were evaluated in pot bioassays and reduced disease incidence by 50%–73%. Strain S3 was as effective at suppressing disease at the lowest inoculum density (106 CFU/mL) as at higher density (108 CFU/mL or undiluted suspension). This study indicates that selected endophytic bacterial strains have potential for control of seedling disease of rice and for plant growth promotion. Key words: biological control, plant growth promotion, endophytes, rice, seedling disease. Résumé : Parmi 102 bactéries endophytes et du rhizoplan isolées à partir de racines et de tiges de riz en Californie, 37% on inhibé significativement (P ≤ 0,05) la croissance in vitro de deux pathogènes, Achlya klebsiana et Pythium spinosum, causant la maladies des semis chez le riz. Quatre souches endophytiques se sont avérées particulièrement efficaces contre la maladie des semis selon des analyses de croissance en pochettes. Le séquençage des gènes amplifiés de l’ARNr 16S nous a permis d’identifier celles-ci comme étant Pseudomonas fluorescens (S3), Pseudomonas tolaasii (S20), Pseudomonas veronii (S21) et Sphingomonas trueperi (S12). Les souches S12, S20 et S21 étaient porteuses du gène de la fixation de l’azote, nifD, mais seule la souche S12 fut capable de réduire l’acétyène en culture pure. Les quatre souches ont significativement augmenté la croissance végétale en l’absence de pathogènes, comme l’ont indiqué des augmentations dans la hauteur des plantes et dans le poids sec de semis de riz inoculés par rapport à du riz non inoculé. Trois souches bactériennes (S3, S20 et S21) ont été soumises à des bioessais en pots et ont diminué de 50%– 73% l’incidence de la maladie. La souche S3 fut en mesure de supprimer efficacement la maladie à la densité d’inoculation la plus faible (106 CFU/mL) aussi bien qu’à une densité plus élevée (108 CFU/mL ou suspension non diluée). Cette étude indique que des souches de bactéries endophytes sélectionnées démontrent un potentiel dans le contrôle de la maladie des semis et dans la stimulation de la croissance végétale. Mots clés : contrôle biologique, stimulation de la croissance des plantes, endophytes, riz, maladies des semis. [Traduit par la Rédaction]

924 Adhikari et al.

Received April 3, 2001. Revision received August 8, 2001. Accepted August 15, 2001. Published on the NRC Research Press Web site at http://cjm.nrc.ca on October 5, 2001. T.B. Adhikari1 and L.M. Nelson.2,3 Agrium Inc., 104-111 Research Drive, Saskatoon, SK S7N 3R2, Canada. C.M. Joseph, G. Yang,4 and D.A. Phillips. Department of Agronomy and Range Science, University of California, Davis, CA 95616, U.S.A. 1

Present address: USDA/ARS, Department of Botany and Plant Pathology, 1155 Lilly Hall, Purdue University, West Lafayette, IN 47907, U.S.A. 2 Corresponding author (e-mail: [email protected]). 3 Present address: Department of Applied Microbiology and Food Science, University of Saskatchewan, Agricultural Buildings, 57 Campus Drive, Saskatoon, SK S7N 5A8, Canada. 4 Present address: MicroBio RhizoGen Corporation, 3835 Thatcher Ave., Saskatoon, SK S7R 1A3, Canada. Can. J. Microbiol. 47: 916–924 (2001)

DOI: 10.1139/cjm-47-10-916

© 2001 NRC Canada

Adhikari et al.

Introduction Seedling disease of rice (Oryza sativa L.), also known as “seed rot”, “water-mold”, and “damping-off”, is a worldwide problem (Ou 1985; Rush 1992). This disease is mainly caused by species of Achlya and Pythium that are most frequently isolated from water-sown rice (Ou 1985; Rush 1992; Webster et al. 1973). When seeds are attacked immediately after sowing, they do not germinate or they fail to emerge through the floodwater because the endosperm or embryo may be infected by species of Achlya and Pythium (Chun and Schneider 1998; Rush 1992). Root infections occur mainly on fine rootlets that are difficult to recover from soil (Chun and Schneider 1998). Aboveground symptoms include damping-off, stunting, and chlorosis. The disease usually develops when seedlings lack vigour and is often exacerbated by cool, cloudy weather that retards germination and seedling growth (Rush 1992). In severe cases, fields must be replanted, which may result in delayed harvest and reduced yields (Webster et al. 1970). Traditionally, chemical and cultural methods have been used to combat seedling disease of rice. Cultural control practices involve precision land leveling to maintain a uniform shallow flood combined with draining and reflooding shortly after planting. However, this practice is costly and alters water and fertilizer nitrogen management (Chun and Schneider 1998). Chemical treatment of rice seeds has been used to control water-mold pathogens of rice (Rush 1992; Rush and Gifford 1972; Webster et al. 1973). However, the application of fungicides in water-sown rice is expensive and not always effective. In addition, there are concerns for health, safety, and environmental risks. Therefore, exploration of an alternative strategy such as biocontrol agents for controlling seedling disease of rice is warranted. A diverse group of introduced microorganisms appears to have potential for biological control of soilborne diseases of rice. Strains of fluorescent and nonfluorescent pseudomonads were effective for the control of sheath blight and bakanae diseases of rice (Mew and Rosales 1986; Rosales et al. 1986). However, there are no reports on antagonism of such bacteria against species of Pythium and Achyla and their potential use as biocontrol agents of seedling disease of rice. The main objectives of this study were to isolate rhizoplane and endophytic bacteria associated with rice in California and to evaluate their efficacy in controlling seedling disease of rice under controlled conditions. As recent studies have demonstrated, the occurrence of nitrogen-fixing endophytic bacteria in rice (Stoltzfus et al. 1997), the presence of the nitrogen fixation gene, nifD, in these bacteria was determined by polymerase chain reaction (PCR) assay.

Materials and methods Isolation and origin of bacterial strains Bacteria isolated from rice roots and stems were collected during May and September 1997 from randomly selected rice fields at Maxwell, Calif., Geer Ranch (Knights Landing, Calif.), and Lundberg Family Farms (Richvale, Calif.). All plants were free of obvious symptoms of disease. Roots and stems were kept on ice until they were processed.

917 Roots were cut from stems, washed with sterile distilled water, and placed in sterile tubes containing sterile water and 0.1% Tween 20 (Sigma Chemical Co., St Louis, Mo.). The tubes were vortexed for 5 min and placed in a bath sonicator (Branson model 1200) for 30 min. Tenfold dilutions of the resulting suspensions were prepared and 100-µL aliquots of the dilutions plated onto tryptic soy agar (TSA), tryptone–yeast extract (TYE) agar (Difco Laboratories, Detroit, Mich.), Vincent’s minimal medium (Vincent 1970), or rice medium supplemented with either nystatin (100 mg/mL) or benomyl (30 mg/mL) and cycloheximide (100 mg/mL). Rice medium contained KH2PO4 (0.5 g/L), K2HPO4 (0.5 g/L), filter sterilized rice liquor (500 mL/L), and 1.5% agar (pH 7). Rice liquor was prepared by adding washed rice stems to 500 mL of deionized water to a total volume of 1 L, blending until the stems were completely macerated, centrifuging (7000 × g, 20 min), and filtering the supernatant through progressively finer filters until a final filter sterilization through 0.22 µm nylon filter was performed. The outer layer of stem samples was removed; the remaining stem core was washed with sterile distilled water and 0.1% Tween 20, cut into 8-cm lengths, placed in sterile beakers, and soaked for 1 h in 1% chloramine T. Stems were rinsed once with 0.1% Tween 20, three times with sterile water, and were cut into lengths of 3– 4 cm. Stem pieces were centrifuged aseptically (12 000 × g, 20 min), stems were immediately removed, and 100-µL aliquots of the remaining liquid were plated directly onto selection media as described above. Plates were incubated at 28°C until colonies appeared. After 5 days, distinct colony types were selected, isolated, and streaked onto half-strength TSA to obtain pure cultures. All 102 strains were stored in 15% glycerol at –80°C until further use. Colony morphology of all bacteria was recorded following growth on half-strength TSA plates incubated at 28°C for 24–48 h.

Bacterial inoculum preparation Bacteria were grown in 250-mL Erlenmeyer flasks containing 100 mL of sterilized tryptic soy broth (TSB) (Difco) on a shaker (Lab-Line Instruments Inc., Melrose Park, Ill.) at 150 rpm for 48 h. Bacterial cells were harvested by centrifugation (12 000 × g, 20°C, 10 min), and the pellet was suspended in 10 mL of sterile distilled water. A dilution series was made of each suspension, and 0.1 mL each of the 10–4, 10–5, and 10–6 dilutions was spread onto TSA plates. Plates were incubated at 28°C for 48 h before determining the colony forming units (CFU) of bacteria per millilitre (CFU/mL). Unless stated otherwise, a concentration of approximately 108 CFU/mL (OD595 = 0.3) of bacteria was used as inoculum.

Fungal inoculum preparation Achlya and Pythium were isolated from rice plants in California by plating roots onto potato dextrose agar (PDA) (Difco) supplemented with neomycin (100 mg/mL) and used in the initial screening of 102 bacterial strains for in vitro antagonism. Pure cultures of Achlya klebsiana Pieters (ATCC 52605) and Pythium spinosum Sawada et Sawada (ATCC 96205) were obtained from American Type Culture Collection (ATCC, Manassas, Va.), grown on PDA plates, and maintained at 4°C. The ATCC strains were used to determine in vitro antagonism by 37 selected bacterial strains, for growth pouch and growth chamber bioassays. For inoculum preparation, ATCC strains of A. klebsiana and P. spinosum were grown separately on PDA plates for 2–3 days and suspended in 120 mL of sterile distilled water. Inoculum was prepared by macerating the entire mycelial mass from a PDA plate in a blender for 30 s at high speed. Optical densities of A. klebsiana (OD600 = 1.44) and P. spinosum (OD600 = 1.48) were recorded prior to inoculation. © 2001 NRC Canada

918

Can. J. Microbiol. Vol. 47, 2001

Rice cultivar A commercial, early-maturing variety of rice (Oryza sativa L.), M202, grown in California, was obtained from the University of California at Davis and used in growth pouch and pot studies.

In vitro assay Bacteria were grown overnight in half-strength TSB and diluted to an OD600 of 0.4–0.7 before plating 100-µL aliquots onto halfstrength TSA plates for fungal biocontrol assays. In preliminary tests of the 102 bacterial strains, four fungal hyphal plugs (two of field-collected isolates of Achlya and two of Pythium) were placed on each plate. In secondary tests, a single hyphal plug was placed on each plate. The inhibitory effect of each bacterial strain on the fungal field isolates was scored by measuring the diameter of fungal growth around the initial hyphal inoculum 24 and 48 h after plating, compared with control plates without bacteria. Of the 102 strains tested, 37 strains inhibited growth of field-collected isolates of Achlya and Pythium and were selected for further studies. For in vitro antagonism assays with the 37 selected strains, agar plugs (3 mm in diameter) from 3-day-old cultures of the ATCC strains of A. klebsiana and P. spinosum grown on PDA were used as inoculum. One loopful (10 µL) of each bacterial suspension was streaked 5 cm in length and 3 cm away from the agar plug placed in the center of the PDA plate. Control plates were not inoculated with bacteria. Plates were kept upright in a plastic container and incubated at 28°C. Assays were conducted twice with three replications each. Growth inhibition was recorded after 2–3 days by the formula of Skidmore (1976):

[1]

Kr – rl/Kr × 100 = GI

where Kr is the distance (measured in mm) of fungal growth from the point of inoculation to the colony margin on control plates, rl the distance of fungal growth from the point of inoculation to the colony margin in the direction of the bacteria, and GI the percent growth inhibition. Percent growth inhibition was categorized on a scale from 0 to 4, where 0 indicates no growth inhibition; 1, 1%– 25% growth inhibition; 2, 26%–50% growth inhibition; 3, 51%– 75% growth inhibition; and 4, 76%–100% growth inhibition.

Identification of bacteria Four bacterial strains (S3, S12, S20, and S21) isolated from rice stems were selected for further study because they were strongly inhibitory to both rice pathogens in the in vitro assay and three of the four promoted rice growth in preliminary growth pouch assays (data not shown). They were identified using two approaches. Whole cell fatty acids were derivatized to methyl esters (FAME) and analyzed by gas chromatography using the MIDI system (Microbial Identification System Inc., Newark, Del.), as described by De Freitas et al. (1997). The FAME profile of Xanthomonas maltophilia ATCC 13637 was used as a reference. The assays were performed twice. These bacterial strains were also identified by sequencing 16S rRNA genes amplified by PCR with primers fD1 and rD2 (Wang et al. 1999).

PCR analysis of nifD gene Two universal nifD primers, FdB260 and FdB261 (Stoltzfus et al. 1997), were used to test for the presence of the nifD gene in all 102 strains. The thermocycler (PTC-100, MJ-Research Inc., Waltham, Mass.) program consisted of three cycles of denaturing (1 min, 94°C), annealing (1 min, 37°C), and extension (2 min, 72°C), followed by 32 cycles of denaturing (1 min, 94°C), annealing (1 min, 55°C), and extension (2 min, 72°C). A 10 µL reaction mixture was electrophoresed on a 1.5% agarose gel using a 1 kB DNA marker ladder (GibcoBRL, Gaithersburg, Md.) as a standard. Control DNA from Sinorhizobium meliloti 1021 produced a 390-bp

fragment in electrophoretic tests as did all strains recorded as being positive in this study.

Acetylene reduction assay The nitrogenase activity of bacterial strains S3, S12, S20, and S21 in pure culture and in the rhizosphere of rice seedlings was determined by the acetylene reduction assay. Twenty-four hour bacterial cultures from half-strength TSB (Difco) were centrifuged at 4000 × g for 5 min and resuspended in 0.1 M MgSO4·7H2O. The population sizes ranged from 6 × 109 (OD660 0.81) to 1.6 × 1010 (OD660 0.86) CFU/mL. Rice seed (cv. M202) was surface sterilized by immersion in 95% ethanol for 20 s followed by a 20% solution of commercial bleach (Clorox Co., Oakland, Calif.) for 10 min and seven rinses in sterile distilled H2O. Seed was germinated by incubation on moist sterile fluted filter paper in the dark at 23°C for 4 days before use. Acetylene reduction assays were conducted in test tubes containing 10 mL of semisolid nitrogen-free medium (Cote and Gherna 1994) with 5.0 g/L of sodium malate for treatments with pure cultures of bacteria and without a carbon source for treatments with a rice seedling. Twenty microlitres of each bacterial culture was inoculated onto the surface of the agar and stabbed 2 cm below the surface with a sterile inoculating loop. In treatments with a plant, a rice seedling was placed on the surface of the agar following bacterial inoculation, and the lower half of the test tube was covered with aluminum foil. The root grew down into the agar during the assay. Noninoculated tubes with and without seedlings served as controls. Test tubes were closed with serum stoppers and 5% of the headspace gas was replaced with purified C2H2 (Matheson Co., Ont., Canada). The treatments were incubated in a controlled environment growth chamber with a day and night temperature of 24°C and 20°C, respectively. Ethylene was measured after 48 h using a gas chromatograph equipped with a Flame Ionization Detector (Nelson and Child 1981). There were three replicate tubes per treatment, and experiments were conducted twice.

Growth pouch assay The effects of bacterial strains S3, S12, S20, and S21 on growth of rice seedlings were assessed in water (control), Hoagland’s solution, and pathogen-infested water using growth pouches (VWR, Mega International, Minneapolis, Minn., U.S.A.). In the first experiment, the growth of rice seedlings in distilled water or in Hoagland’s solution was compared. The pouches were wetted with 10 mL of distilled water or 0.25× modified Hoagland’s nitrogen free solution (Hoagland and Arnon 1938), containing 0.1 mM KH2PO4, 2 mM K2SO4, 0.1 mM CaSO4, 2 mM MgSO4, 0.5 mM iron citrate, and 1 mL of micronutrient solution. The micronutrient solution contained (per litre) 1 g H3BO3, 1 g MnCl2·4H20, 58 g ZnSO4·7H2O, 0.13 g CuSO4·5H2O, and 0.10 g Na2MoO4·2H2O) at pH 6.8 and was filter sterilized with a 0.22-µm filter (VWR Canlab, Mississauga, Ont., Canada). In the second experiment, the effect of the bacterial strains on growth of rice seedlings in the presence of two pathogens was determined in pouches containing 10 mL of distilled water. A 1-mL suspension of A. klebsiana (OD600 = 1.44) or P. spinosum (OD600 = 1.48) was added to each pouch. Control pouches were not infested with pathogens. Rice seeds (5 g) were surface sterilized for 30 min in a freshly prepared 10% solution of commercial bleach (Clorox) and were rinsed four times in 250 mL of sterile distilled water. Seeds were dried in a laminar flow cabinet overnight before planting. Inocula of bacterial strains (S3, S12, S20, and S21) were prepared as described above, and rice seeds were soaked for 1 h in each bacterial suspension (108 CFU/mL). After soaking, the excess broth was drained, and the seeds were allowed to dry for 2 h in a laminar flow cabinet. Pouches were then planted with the surface sterilized rice seeds (five seeds per pouch), placed in holding racks, and covered with © 2001 NRC Canada

Adhikari et al.

919

Table 1. Identity of strains using FAME analysis or 16S rRNA gene sequencing; growth inhibition of Achlya klebsiana and Pythium spinosum in vitro; acetylene reduction activity (ARA) in pure culture (PC) and in rice rhizosphere (RR); and detection of amplified nifD fragment in bacterial strains isolated from rice stems in California.

Strain S3 S12 S20 S21

FAME (Similarity index) Pseudomonas putida Biotype B (0.77) No match Pseudomonas putida Biotype A (0.41) Pseudomonas putida Biotype A (0.29)

Growth inhibition (GI) category*

ARA†

16S rRNA (% identity)

A. klebsiana

P. spinosum

PC

RR

nifD‡

Pseudomonas fluorescens (99.6) Sphingomonas trueperi (99.8) Pseudomonas tolaasii (99.6) Pseudomonas veronii (99.8)

3

3

–

–

–

1

2

+

–

+

2

2

–

–

+

2

2

–

–

+

*Values on a scale from 0 to 4: 0, no growth inhibition; 1, 1%–25%; 2, 26%–50%; 3, 51%–75%; and 4, 76%–100%. † +/– indicates presence/absence of ethylene following incubation for 48 h of each strain under nitrogen-free conditions in pure culture or in the rhizosphere of 4-day-old rice seedlings. ‡ +/– indicates the presence/absence of the nifD fragment following PCR amplification.

plastic wrap. The pouches were incubated at 28°C with a 12-h photoperiod. Rice seeds soaked in sterile distilled water served as controls. Assays were conducted with three replicates of 25 seeds per replicate for each treatment. After 10 days, root and shoot lengths of seedlings were assessed.

Disease bioassay in growth chamber Two strains, S20 and S21, were further evaluated in a growth chamber. Rice seeds were inoculated with bacteria, as described above. PDA plugs (0.7 cm diameter) of A. klebsiana and P. spinosum were inoculated into sterile culture bottles containing 300 g of wheat seed in 100 mL of H2O and were incubated at 20°C for 2 weeks. Bioassays were conducted in 10 cm diameter Jiffy Pots (W. R. Grace and Co. Canada Ltd., Ajax, Ont.), each containing 150 g of nonsterilized potting mix #3 (Sun Gro Horticulture Canada Ltd., Ont., Canada). The soil was fertilized at the rate of 120–60–40 N–P–K (5 g of N–P–K per pot) in the forms of urea, single super phosphate, and muriate of potash, prior to inoculation. Each pot soil was infested by mixing 10 g of inoculum of A. klebsiana or P. spinosum just prior to seeding. Four rice seeds (with or without bacterial treatment) per pot were immediately sown in pathogen-infested and noninfested soils (controls). The pots were kept in a growth chamber (model PGR-15, Conviron Winnipeg, Man., Canada) with day:night temperatures of 25°C:22°C and with a 16-h photoperiod (150 µE·m–2·s–1). Pots were watered daily, and 0.03% micronutrient solution (Plant Products Co., Ont., Canada) was added to each pot (20 mL/pot) 2, 4, and 6 weeks after seeding. Disease incidence (%) was calculated on the basis of number of rice seedlings per pot exhibiting typical rooting and damping-off symptoms caused by A. klebsiana or P. spinosum 2–6 weeks after inoculation. To confirm the presence of A. klebsiana or P. spinosum, root systems of infected seedlings were surface disinfested in 0.5% sodium hypochlorite and plated on PDA. Plant height (cm) was recorded 8 weeks after seeding. Plants were harvested from each treatment, oven-dried at 80°C for 48 h, and seedling dry weight (g) was recorded. Each treatment consisted of five replicate pots of four plants per pot and arranged in a randomized complete block design. The experiment was conducted twice.

Effect of inoculum densities of S3 Strain S3 exhibited strong inhibitory effects on major fungal pathogens of rice, including A. klebsiana, P. spinosum, Rhizoctonia solani Kühn, and Magnaporthe grisea (Hebert) Barr (data not

shown), and phytotoxicity was observed in our initial experiments. Therefore, a series of bioassays was designed to test the effect of inoculum density of S3 on phytotoxicity, antagonism, growth promotion, and disease suppression in vitro, in growth pouches and in a growth chamber. Strain S3 was grown in 250 mL Erlenmeyer flasks containing 100 mL of TSB. After shaking for 48 h (100 rpm) at 28°C, cells were harvested by centrifugation at 12 000 × g for 10 min. The resulting pellet was suspended in 25 mL of sterile distilled water, and a dilution series was made to concentrations of 108, 107, 106, and 105 CFU/mL. All experimental procedures for in vitro inhibition, growth pouch, and growth chamber bioassays were similar to those described above. Seed germination was determined by counting the number of germinating seeds in each growth pouch. Percent emergence was determined as the number of emerging seedlings in pathogen-infested soil 2 weeks after inoculation in the growth chamber divided by the total number of seeds planted, multiplied by 100. Root and shoot length, root and shoot fresh weight, plant height, seedling dry weight, and disease incidence were assayed, as described above. The undiluted bacterial suspension (5 × 109 CFU/mL) and all concentrations of S3 were evaluated in the presence of A. klebsiana or P. spinosum. A control with a pathogen-infested soil and no bacterial inoculation was included. All experiments had four or five replications of each treatment and were conducted two or three times.

Statistical analyses All experiments were arranged in a completely randomized block design. In growth chamber studies, experiments were repeated and all data pooled, provided there were no interactions between experiments and treatments. Data were subjected to analysis of variance (ANOVA) and treatment means compared using Duncan’s Multiple Range Test (DMRT) at P ≤ 0.05. The Statistical Analysis Systems (SAS) was used for all analyses (SAS Institute Inc. 1989).

Results Identification of bacterial strains Of the 102 rhizoplane and endophytic bacteria isolated from rice roots and shoots, 37% were selected for further study because they were antagonistic to field-collected isolates of Achlya and Pythium spp. from California or they © 2001 NRC Canada

920

Can. J. Microbiol. Vol. 47, 2001

Table 2. Effect of seed inoculation with four bacterial strains on root and shoot length of rice seedlings in growth pouch assays.

Treatment Control* S3 S12 S20 S21

Root length (cm)

Shoot length (cm)

Water

Water

13.2d 14.1c 14.9ab 15.3a 14.5b

Hoagland 13.6b 15.4a 15.1a 15.4a 15.3a

3.0c 3.3b 3.3b 4.0a 3.9a

Table 3. Effect of seed inoculation with four bacterial strains on root and shoot length of rice seedlings in growth pouches infested with Achlya klebsiana or Pythium spinosum. Root length (cm)

Shoot length (cm)

Treatment

Pythium

Achlya

Pythium

Achlya

Control* S3 S12 S20 S21

4.7c 6.2bc 7.3ab 8.5a 8.6a

11.6c 12.5bc 14.2a 13.2ab 13.2ab

1.8b 2.6a 2.1ab 2.4a 2.2ab

2.1c 3.0b 4.0a 3.2b 3.2b

Hoagland 3.7b 4.7a 4.4a 4.6a 4.6a

Note: Root and shoot length were assessed 10 days after inoculation. Values are the means of three replications (25 seeds/replicate). Means within columns followed by same letter are not significantly different (P ≤ 0.05) by Duncan’s Multiple Range Test. *Noninoculated rice seeds grown in water or 0.25× modified Hoagland’s solution.

amplified a nifD-gene fragment (see below). A subset of four strains S3, S12, S20, and S21, all putative endophytes from rice stems, were evaluated and retained for further studies because they were the most inhibitory, ranging from 25% to 75% to both A. klebsiana (ATCC 52605) and P. spinosum (ATCC 96205) (Table 1). One strain, S3 was extremely inhibitory and had the highest GI category score against both rice pathogens. Three of the four strains were identified as pseudomonads by FAME and 16S rRNA methods (Table 1). Stronger similarity to known strains was obtained with the 16S rRNA sequence method. Strain S12, which did not match any strain in the MIDI database, was identified as Sphingomonas trueperi by 16S rRNA gene analysis. Colonies on TSA medium ranged from small to medium with off-white (S3, S20, and S21) or yellow (S12) color at 28°C. Nitrogen fixation capacity In PCR assays, only five of the initial 102 isolates reproducibly amplified appropriate products with the nifD gene primers, and four of these were isolated from rice stems at the Geer ranch site. Of these four endophytic strains, three (S12, S20, and S21) also had been selected based on their ability to inhibit growth of field-collected isolates of Achlya and Pythium (Table 1). Although these three strains were positive for nifD, only S12 demonstrated acetylene reduction activity in nitrogen-free malate medium and none showed any acetylene reduction activity in the rhizosphere of rice seedlings (Table 1). Growth pouch assay Rice seedling vigour, as measured by root and (or) shoot length in water or Hoagland’s nitrogen-free solution, was significantly increased by inoculation with bacteria compared with that of the noninoculated control (P ≤ 0.05) (Table 2). ANOVA showed that Hoagland’s nitrogen-free solution supported significantly better root and shoot development than water (data not shown, P ≤ 0.05). In pathogeninfested water, roots and shoots were significantly longer for rice inoculated with bacteria than for noninoculated controls (P ≤ 0.05) (Table 3). ANOVA showed that root and shoot lengths in plants infected with P. spinosum were significantly shorter than those in plants infected with A. klebsiana (data not shown, P ≤ 0.05).

Note: Root and shoot lengths were measured 10 days after inoculation. Values are the means of three replications (25 seeds/replicate). Means within columns followed by same letter are not significantly different (P ≤ 0.05) by Duncan’s Multiple Range Test. *Noninoculated rice seeds grown in water infested with pathogen.

Disease bioassay in growth chamber Inoculation with either S20 or S21 significantly increased plant height (4%–10%) and seedling dry weight (12%–50%) compared with noninoculated controls in pathogen-infested soil (Figs. 1A and 1B) (P ≤ 0.05). Both S20 and S21 significantly reduced disease incidence (50%–73%) compared with noninoculated plants in pathogen-infested soil (Fig. 1C) (P ≤ 0.05). Effects of strain S3 inoculum densities on bioassays The growth of ATCC strains of A. klebsiana and P. spinosum was significantly inhibited in vitro by strain S3 (P ≤ 0.05), but varied with inoculum density (Fig. 2A). Regardless of inoculum density, antagonism to A. klebsiana was greater than to P. spinosum, but no significant differences were observed among undiluted suspensions, 108, 107, and 106 CFU/mL. Antagonism to P. spinosum differed significantly and decreased gradually with decreasing inoculum density. Rice seed treated with inoculum densities at 108, 107, 106, and 105 CFU/mL had more than 90% seed germination, whereas application of undiluted suspension of S3 resulted in significantly lower seed germination to 42% (Fig. 2B) (P ≤ 0.05). Application of S3 inoculum densities ranging from 105 to 108 CFU/mL significantly increased root and shoot length (Fig. 3A) and fresh weight (Fig. 3B) of rice seedlings in growth pouches in the absence of pathogens compared with controls (P ≤ 0.05). Compared with other inoculum densities, root and shoot length and root and shoot fresh weight were significantly reduced by the undiluted suspension of S3. S3 applied at inoculum densities of 105–106 CFU/mL increased emergence in pathogen-infested soil relative to noninoculated controls (Fig. 4) (P ≤ 0.05). Treatment of rice seed with S3 at each of the inoculum densities significantly enhanced plant height in pots in the presence of Pythium (P ≤ 0.05), but no differences were observed among inoculum densities (Fig. 5A). Inoculation of rice seed with S3 significantly increased plant dry weight by 16%–50% compared with noninoculated controls in the presence of Pythium (Fig. 5B) (P ≤ 0.05). In soils infested with Achlya inoculum densities of 108 and 105 CFU/mL were effective in enhancing plant height and dry weight compared with controls (Figs. 5A and 5B). Although inocula© 2001 NRC Canada

Adhikari et al. Fig. 1. Effects of bacterial strains S20 and S21 on plant height (A), plant dry weight per pot (B), and disease incidence (C), in a growth chamber. Rice seeds were inoculated with each bacterial strain and noninoculated seeds served as the control. Plant height (cm) and dry weight (g) were recorded 8 weeks after sowing. Disease incidence (%) is represented by the number of seedlings infected divided by the total seedlings in each treatment and multiplied by 100. Each treatment consisted of five replicate pots with four seedlings per pot arranged in a completely randomized block design. Values are the means of two experiments. Means with the same letter are not significantly different from each other, according to Duncan’s Multiple Range Test (P ≤ 0.05).

tion of rice seed with S3 at densities from 105 to 108 CFU/mL or with undiluted suspension reduced disease incidence in both pathogen-infested soils by approximately 50%, no significant differences were observed among inoculum densities (Fig. 5C) (P ≤ 0.05).

Discussion This study demonstrates that bacterial strains with the potential for biological control of seedling disease of rice can be selected from rice stems. When rice seed was inoculated with selected strains (S3, S12, S20, or S21) and infected with A. klebsiana or P. spinosum, significant increases in seedling height and dry weight compared with noninoculated plants were the most consistent indications of dis-

921 Fig. 2. Effects of inoculum densities of S3 on in vitro antagonism to Achlya klebsiana and Pythium spinosum (A) and on seed germination (B), in a growth pouch assay. In in vitro antagonism assays, percent growth inhibition was determined 2–3 days after inoculation, according to the formula of Skidmore (20). Values are the means of three replicates, and noninoculated plates with bacteria served as controls. In growth pouch assays, rice seeds were inoculated with S3 at each inoculum density, and noninoculated seeds served as the control. Percent germination was determined 10 days after inoculation. Each treatment consisted of four replicates with 25 seeds/replicate. Values are the means of two experiments. Means with the same letter are not significantly different from each other, according to Duncan’s Multiple Range Test (P ≤ 0.05).

ease suppression. Three of the four strains selected for disease suppression also contained DNA that produced appropriate sized PCR products with nifD primers. However, one strain, S3 that showed no evidence for nifD, demonstrated the greatest potential for biological control of the two fungal pathogens of rice. Classification of the four selected strains varied depending on the approach used, but both FAME and 16S rRNA identified S3, S20, and S21 as belonging to the same genus, Pseudomonas. Other bacterial strains of this genus have demonstrated potential for biological control of fungal pathogens of rice (Mew and Rosales 1986; Rosales et al. 1986). Similarity indices for S20 and S21 by FAME analysis were less than 0.5, indicating poor matches. The amplification and comparison of 16S rRNA sequences is a sensitive tool for identification of bacteria (Weisburg et al. 1991) and provided highly similar sequence identities to known species © 2001 NRC Canada

922 Fig. 3. Effects of inoculum densities of S3 on root and shoot length (A) and on root and shoot fresh weight (B), in growth pouch assays without pathogens. Rice seeds were inoculated with each bacterial strain, and noninoculated seeds served as the control. Root and shoot length and root and shoot fresh weight were determined 10 days after inoculation. Values are the means of four replicates with 25 seeds/replicate. Means with the same letter are not significantly different from each other, according to Duncan’s Multiple Range Test (P ≤ 0.05).

for the four strains studied. S12 was identified as Sphingomonas, and this genus is commonly found in soil and water and also on plant surfaces (Kim et al. 1998). It was the only nifD-positive isolate able to fix nitrogen in pure culture. Sphingomonas trueperi was originally reported as a nitrogen-fixing bacterium (Anderson 1955), but subsequent workers found no evidence for nitrogenase activity (Hill and Postgate 1969). Nitrogen-fixing isolates of Sphingomonas paucimobilis have been identified from the rhizosphere of rice (Bally et al. 1990). Stoltzfus et al. (1997) reported that only 13 of 17 nifD positive endophytic isolates from rice reduced acetylene. The conditions required for expression of nitrogen fixation may not have been supplied in the nitrogen free malate medium for S20 and S21 or in the seedling rhizosphere for the three nifD-positive strains reported here. As nifD is only one of at least 20 genes required for nitrogen fixation (Postgate 1998), strains such as S20 and S21 may lack other genes required for nitrogenase function. The inoculation of rice seeds with endophytic bacteria and addition of nitrogen free Hoagland’s solution enhanced seedling development in growth pouch assays relative to noninoculated water controls in disease-free conditions. This enhancement of seedling vigour may result from rapid multiplication of bacteria in the root and stem systems in re-

Can. J. Microbiol. Vol. 47, 2001 Fig. 4. Effects of inoculum densities of S3 on rice plants grown in pathogen-infested soil in a growth chamber. Rice seeds were inoculated with each bacterial strain, and noninoculated seeds served as the control. Percent emergence was calculated as the number of seedlings that emerged in pathogen-infested soil 2 weeks after inoculation divided by the total number of seeds planted multiplied by 100. Each treatment consisted of five replicate pots with four seedlings per pot arranged in a completely randomized block design. Values are the means of two experiments. Means with the same letter are not significantly different from each other, according to Duncan’s Multiple Range Test (P 0.05).

sponse to root exudates and from the production by these bacteria of beneficial substances such as plant growth regulators (Glick 1995). It has been suggested that microorganisms isolated from the roots or shoots of a specific crop may be better adapted to that crop and may provide better disease control than organisms originally isolated from other plant species (Weller 1988; Cook 1993). Such plant-associated microorganisms may make better biocontrol agents because they are already closely associated with the plant as well as being adapted to the local environment. The strains used in this study were isolated from the roots and stems of healthy rice in California and were apparently well adapted to the rice rhizosphere. The endophytic bacteria tested in this study were selected initially on the basis of in vitro antagonism to A. klebsiana and P. spinosum. Although antagonism to fungal pathogens in vitro is not necessarily correlated with in vivo activity (Burr et al. 1978), identification of in vitro antibiosis can be useful in the selection of effective strains (Weller and Cook 1986). All four strains (S3, S12, S20, and S21) provided significant disease suppression compared with noninoculated control seeds when rice seedlings were challenged with the two water-mold pathogens individually. Testing of the other isolated strains for in vivo activity may identify other organisms with the potential for biological control of these pathogens. Since the root is a major site of infection for Pythium or Achlya (Chun and Schneider 1998), colonization and protection of roots by a potential biocontrol agent may be important for the control of seedling disease of rice. The beneficial effects of rhizobacteria on plant health management have been demonstrated for several other host–pathogen systems (Burr et al. 1978; Mew and Rosales 1986; Rosales et al. 1986; Sakthivel et al. 1986; Suslow and © 2001 NRC Canada

Adhikari et al. Fig. 5. Effects of inoculum densities of S3 on plant height (A), plant dry weight (B), and disease incidence (C), in a growth chamber. Rice seeds were inoculated with each bacterial strain and noninoculated seeds served as the control. Plant height and dry weight were recorded 8 weeks after sowing. Disease incidence (%) represents the number of infected seedlings divided by the total number of seedlings in each treatment and multiplied by 100. Each treatment consisted of five replicate pots of four seedlings per pot arranged in a completely randomized block design. Values are the means of two experiments. Means with the same letter are not significantly different from each other, according to Duncan’s Multiple Range Test (P ≤0.05).

923

plant growth under controlled conditions. Of the four strains evaluated in this study, S3 appeared to be the most promising biocontrol agent for further study. It was generally an effective suppressor at low inoculum density (106 CFU/mL) of both A. klebsiana and P. spinosum and promoted rice growth. Further experiments are needed to determine the effectiveness of these isolates under field conditions and to understand the nature of their interaction with the pathogens and the host plant.

Acknowledgements We thank National Sciences and Engineering Research Council (NSERC) and Canada and Agrium Inc. for providing financial support, Dr. J. Germida for fatty acid analysis, and K. McCaffery, M. Dworazcek, and Dr. R. Hynes for conducting acetylene reduction assays.

References

Schroth 1982; Weller and Cook 1983, 1986). Although many strains have shown antibiosis as a mechanism for biological control (Cook 1993; Weller 1988; Weller and Cook 1983), further work is needed to verify the antagonistic properties of these rice endophytic bacteria in suppressing the activities of pathogenic organisms in the rhizosphere. The inoculum density of bacteria influenced the magnitude of biological control and rice growth. With P. fluorescens S3, phytotoxicity was evident in initial experiments, whereas no phytotoxicity was observed when rice seeds were treated with strains S20 and S21. Subsequent experiments indicated that the phytotoxicity occurred only with the highest inoculum densities (above 107 CFU/mL) of S3, although antagonism towards A. klebsiana and P. spinosum was maximum at the higher inoculum densities. This study provided an initial assessment of the potential of rhizoplane and endophytic bacteria associated with rice in California to control seedling disease of rice and promote

Anderson, G.R. 1955. Nitrogen fixation by Pseudomonas-like soil bacteria. J. Bacteriol. 70: 129–133. Bally, R., Givaudan, A., Bernillon, J., Heulin, T., Balandreau, J., and Bardin, R. 1990. Numerical taxonomic study of three N2fixing yellow-pigmented bacteria related to Pseudomonas paucimobilis. Can. J. Microbiol. 36: 850–855. Burr, T.J., Schroth, M.N., and Suslow, T. 1978. Increased potato yields by treatments of seedpieces with specific strains of Pseudomonas fluorescens and P. putida. Phytopathology, 68: 1377–1383. Chun, S.-C., and Schneider, R.W. 1998. Sites infection by Pythium species in rice seedlings and effects of plant age and water depth on disease development. Phytopathology, 88: 1255–1261. Cook, R.J. 1993. Making greater use of introduced microorganisms for biological control of plant pathogens. Annu. Rev. Phytopathol. 31: 53–80. Cote, R.J., and Gherna, R.L. 1994. Nutrition and media. In Methods for general and molecular bacteriology. Edited by P. Gerhardt, R.G.E. Murray, W.A. Wood, and N.R. Krieg. American Society of Microbiology, Washington, D.C. pp. 155–178. De Freitas, J.R., Bannerjee, M.R., and Germida, J.J. 1997. Phosphate-solubilizing rhizobacteria enhance the growth and yield but not phosphorus uptake of canola (Brassica napus L.) Biol. Fertil. Soils, 24: 358–364. Glick, B.R. 1995. The enhancement of plant growth by free-living bacteria. Can. J. Microbiol. 41: 109–117. Hill, S., and Postgate, J.R. 1969. Failure of putative nitrogen-fixing bacteria to fix nitrogen. J. Gen. Microbiol. 58: 277–285. Hoagland, D.R., and Arnon, D.R. 1938. The water culture method for growing plants without soil. Calif. Agric. Exp. Stn. Bull. 347: 36–39. Kim, H., Nishiyama, M., Kunito, T., Senoo, K., Kawahara, K., Murakami, K., and Oyaizu, H. 1998. High population of Sphingomonas species on plant surface. J. Appl. Microbiol. 85: 731–736. Mew, T.W., and Rosales, A.M. 1986. Bacterization of rice plants for control of sheath blight caused by Rhizoctonia solani. Phytopathology, 76: 1260–1264. Nelson, L.M., and Child, J.J. 1981. Nitrogen fixation and hydrogen metabolism by Rhizobium leguminosarum isolates in pea root nodules. Can. J. Microbiol. 27: 1028–1034. Ou, S.H. 1985. Rice diseases. Commonwealth Mycological Institute, Kew, England. © 2001 NRC Canada

924 Postgate, J. 1998. Nitrogen fixation. Cambridge University Press, Cambridge, U.K. Rosales, A.M., Nuque, F.L., and Mew, T.W. 1986. Biological control of bakanae disease of rice with antagonistic bacteria. Philipp. Phytopath. 22: 29–35. Rush, M.C. 1992. Seedling diseases. In Compendium of rice diseases. Edited by R.K. Webster and P.S. Gunnell. The American Phytopathological Society, St. Paul, Minn. pp. 12–13. Rush, M.C., and Gifford, J.R. 1972. Enhancement of fungicidal control of seed rot and seedling diseases of rice by an insecticidal seed treatment. Plant Dis. Rep. 56: 154–157. Sakthivel, N., Sivamani, E., Unnamalai, N., and Gnanamanickam, S. 1986. Plant growth-promoting rhizobacteria in enhancing plant growth and suppressing plant pathogens. Curr. Sci. 55: 22–25. SAS Institute Inc. 1989. SAS/STAT® user’s guide, version 6. 4th ed. SAS Institute Inc., Cary, N.C. Skidmore, A.M. 1976. Interactions in relation to biological control of plant pathogens. In Microbiology of aerial plant surfaces. Edited by C.H. Dickinson and T.F. Preece. Academic Press, London. pp. 507–528. Stoltzfus, J.R., So, R., Malarvithi, P.P., Ladha, J.K., and de Bruijn, F.J. 1997. Isolation of endophytic bacteria from rice and assessment of their potential for supplying rice with biologically fixed nitrogen. Plant Soil, 194: 25–36. Suslow, T.V., and Schroth, M.N. 1982. Rhizobacteria of sugar beet: effects of seed application and root colonization on yield. Phytopathology, 72: 199–206.

Can. J. Microbiol. Vol. 47, 2001 Vincent, J.M. 1970. A manual for the practical study of rootnodule bacteria. I.B.P. Handb. No. 15. Blackwell Scientific Pub. Ltd., Oxford, U.K. Wang, E.T., van Berkum, P., Sui, X.H., Beyene, D., Chen, W.X., and Martinex-Romero, E. 1999. Diversity of rhizobia associated with Amorpha fruticosa isolated from chinese soils and description of Mesorhizobium amorphae sp.nov. Int. J. Syst. Bacteriol. 49: 51–65. Webster, R.K., Hall, D.H., Heeres, J., Wick, C.M., and Brandon, D.M. 1970. Achlya klebsiana and Pythium species as primary causes of seed rot and seedling disease of rice in California. Phytopathology, 60: 964–968. Webster, R.K., Hall, D.H., Bostad, J., Wick, C.M., Brandon, D.M., Baskett, R., and Williams, J.M. 1973. Chemical seed treatment for the control of seedling disease of water-sown rice. Hilgardia, 41: 689–698. Weisburg, W.G., Barns, S.M., Pelletier, D.A., and Lane, D.J. 1991. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 173: 697–703. Weller, D.M. 1988. Biological control of soilborne plant pathogens in the rhizosphere with bacteria. Annu. Rev. Phytopathol. 26: 379–407. Weller, D.M., and Cook, R.J. 1983. Suppression of take-all of wheat by seed treatments with fluorescent pseudomonads. Phytopathology, 73: 463–469. Weller, D.M., and Cook, R.J. 1986. Increased growth of wheat by seed treatments with fluorescent pseudomonads, and implications of Pythium control. Can. J. Plant Pathol. 8: 328–334.

© 2001 NRC Canada