Microb Ecol (1994) 28:101-110

MICROBIAL ECOLOGY © 1994Springer-VerlagNew York Inc.

Epiphytic Bacteria Antagonistic to Curvularia Leaf Spot of Yam S.J. Michereff, N.S.S. Silveira, A. Reis, R.L.R. Mariano Federal Rural University of Pernambuco, Department of Agronomy, 52171-900 Recife, PE, Brazil Received: 2 August 1993; Revised: 31 March 1994

Abstract. Curvularia eragrostidis yam leaf spot is a serious concern among the northeast Brazilian yam growing areas. In order to study its biocontrol, bacterial isolates from the yam phylloplane were tested against the pathogen. They were evaluated with respect to the following parameters: (1) inhibition of C. eragrostidis mycelial growth by using paired culture and cellophane membrane methods, (2) inhibition of conidium germination by using a paired suspension test, (3) reduction of disease severity and, (4) persistence of antagonistic action, on plants under greenhouse conditions. From a total of 162 bacterial isolates, 39 showed antagonism to the pathogen in paired culture. The bacteria produced extracellular, nonvolatile, and diffusible metabolites in the membrane cellophane test. Seventeen isolates resulted in more than 75% inhibition of C. eragrostidis mycelial growth. Among them, IF-26 showed the greatest antagonism. The isolates IF-82, IF-88, and IF-109 inhibited pathogen conidial germination, with average inhibition levels of 99.2, 98.2 and 96.2%, respectively. Under greenhouse conditions the antagonists were applied at three different time intervals relative to C. eragrostidis inoculation: 3 days before, at the same time, and 3 days after. IF-82 and IF-88 applied at the same time as pathogen inoculation both reduced disease severity 75%. IF-82 showed the best persistence of antagonistic action, with an average of 96.3%. IF-82, identified as Bacillus subtilis, was the best biocontrol agent for the yam leaf spot disease in this study. Introduction The leaf spot of yam (Dioscorea cayennensis Lam.), caused by Curvularia eragrostidis (Henn.) Meyer, is one of the most frequent and severe diseases in all yam growing areas of northeast Brazil. The disease causes a reduction of about 35-40% of the weight of the commercial tuber [26]. Disease control usually consists of weekly or biweekly preventative sprays with dithiocarbamate fungicides (manganese ethylene bisditiocarbamate and manganese plus zinc ethylene bisditiocarba-

Correspondence to: Sami J. Michereff

102 Table 1.

S.J. Michereff et al. Characterization of yam leaf sample conditions for isolating bacterial epiphytes Age

Area A B C D E F G H I

Locality Mamanguape, Mamanguape, Mamanguape, Mamanguape, Alhandra, PB Alhandra, PB Goiana, PE Igarassu, PE Recife, PE

(mo) PB PB PB PB

6 4 3 4 5 4 5 6 5

Fungicide (mancozeb) Condition healthy healthy diseased diseased healthy diseased diseased diseased healthy

Applications 0 3 0 0 3 3 3 -

Last application 30 days 5 days 20 days 15 days -

mate) [26]. However there are reports of loss of efficacy for these products (Ramos JEL, MS Thesis, 1991, Federal Rural University of Pernambuco, Brazil). Microbial antagonists have been studied as a means of improving control alternatives [2,10,18]. The use of bacteria as a controlling agent of plant diseases has been shown to be a strategy with great potential [17,22,28,29]. Despite extensive research on the biological control of phylloplane diseases [2, 9], to the present date C. eragrostidis biocontrol has not been studied. This work aims to study the biocontrol potential of yam leaf spot with epiphytic bacteria isolated from the phylloplane.

Materials and Methods

Isolation of Phylloplane Epiphytic Bacteria The bacteria used in this study were isolated from yam leaf (cv. Da Costa) samples taken from nine nurseries in the Paraiba and Pernambuco states, Brazil (Table 1). From each plant, samples with five leaves were taken randomly. A total of 15 plants were sampled per nursery. Five discs (5 ram) were removed per leaf, placed in a tube containing sterile tap water (10 ml), and washed with mechanical shaking for 15 min. Several dilutions were plated on nutrient-yeast extract-dextrose-agar medium (NYDA), which contained beef extract (3g), peptone (5g), glucose (10g), yeast extract (5g), agar (15g), and distilled water (1000 ml). Plates were incubated for 48 h at 25°C, and isolated colonies were picked randomly and purified by streaking on the same medium. Pure cultures were stored at 4°C on nutrient-agar medium (NA).

Laboratory Experiments The ability of the bacterial isolates to produce inhibitory substances against C. eragrostidis was evaluated using the following methods: paired culture, cellophane membrane, and paired suspension. In all experiments C. eragrostidis was grown on potato-dextrose-agar medium (PDA) (cooked potato, 200g; glucose, 20g; agar, 18g; distilled water, 1000 ml) for 8 days, and bacteria were grown on NA medium for 48 h. In the paired culture test, 162 bacterial isolates were evaluated. Two agar disks (5 mm in diameter, 2 mm thick) from cultures of C. eragrostidis were placed 70 mm apart in a Petri dish containing PDA

Antagonism to Curvularia Leaf Spot

103

medium. Plates were incubated 12 h at 25 --- 2°C; each bacterial isolate was streaked as a band in the center of a plate, except for the controls. Thirty-nine bacterial isolates which had shown antagonism in the paired culture test, plus forty-one randomly chosen isolates, were evaluated for the production of inhibitory nonvolatile compounds through the cellophane membrane method. Cellophane membranes (Votacel Inc., S~o Paulo, Brazil) were washed × 3 in tap water and sterilized by autoclaving. A loopfull of bacterial growth was placed centrally on cellophane membranes over PDA in Petri dishes. The membranes were removed after 96 h and an agar disc (5 ram) of C. eragrostidis was placed in the center of each plate. Controls did not receive the bacteria. In both paired and cellophane methods, plates were incubated under alternate light and dark (12 h light/12 h dark) at 25 - 2°C and 70% humidity. They were evaluated after 5 days by measuring the pathogen linear mycelial growth. The percentage of mycelial growth inhibition (MGI) was calculated according to the following formula: MGI (%) = [(MGC - MGT)/MGC] x 100, where MGC = length of mycelial growth on the control, and MGT = length of mycelial growth on the treatment. Seventeen bacterial isolates, which showed antagonism (>75% micelial growth inhibition) in the cellophane test, were studied by the paired suspension method. Droplets (0.2 ml) of sterile distilled water containing C. eragrostidis (2 × l0 s conidia/ml) and the test bacterium (108 cfu/ml) were placed in a cavity of a glass depression slide. The controls did not receive a bacterial suspension. After an incubation period of 6 h, a drop of cotton blue was placed in each glass slide cavity and the number of germinated conidia were counted under a light microscope, with × 100 magnification. A conidium was considered germinated when the length of the germ tube was twice its greatest width. The percentage of germinated conidia was calculated from the average of five fields for each replication. The percentage of conidia germination inhibition (CGI) was calculated according to the following formula: CGI (%) = [(CGC-CGT)/CGC] × 100, where CGC = conidia germination in the control, and CGT = conidia germination in the treatment.

Greenhouse Experiments Seventeen bacterial isolates demonstrating the greatest antagonism by laboratory methods were assessed for their ability to reduce the severity of yam leaf spot and for the persistence of their antagonistic action. Fungal suspensions containing 2 × 10s conidia/ml in distilled water were prepared from 8-day-old cultures grown on PDA. Bacterial suspensions containing 108 cfu/ml in distilled water were prepared using 48 hour-old cultures grown on NA. The suspensions were ammended with 0.05% Tween 80. Greenhouse-grown, 4-month-old yam plants, cv. Da Costa, were inoculated with the C. eragrostidis suspension and treated with bacterial suspension. Sprays were performed until run-off using a De Vilbiss atomizer. Antagonists were applied at three different times in relation to pathogen inoculation: 3 days before, at the same time, and 3 days after. The controls were inoculated with C. eragrostidis but not with bacteria. Following inoculation, plants were placed for 36 h in a moist chamber (31 -+ 2°C and 98.5% humidity) and then returned to greenhouse conditions (32 -+ 2°C and 87% humidity). The disease severity was evaluated 14 and 28 days after inoculation by grading disease symptoms from 0 to 4 (where 0 = without symptoms; 1 = few isolated lesions; 2 = isolated lesions comprising 20% of the leaf area; 3 = many isolated and/or confluent lesions comprising 20-40% of the leaf area; 4 = abundant confluent lesions comprising more than 40% of the leaf area). The percentage of disease severity reduction (DSR) was calculated according to the formula: DSR (%) = [(DSC - DST)/ DSC] x 100, where DSC = disease severity on the control, and DST = disease severity on the treatment. The persistence of the antagonistic action (PAA) for the five most efficient bacterial antagonists was assayed according to the formula: PAA (%) = (DSR28 - DSR14) × 100, where DSR28 = disease severity reduction at 28 days, and DSR14 = disease severity reduction at 14 days.

104

S.J. Michereffet al.

18 ¸ N

u

18

b e r

12

II1

o

f

9

I o !

6

t e

3

8

0 18-20

21-30 31"40 My(:elllll G r o w t h inhibition (%)

41.48

Fig.

1.

Identification of Bacterial Strain IF-82 The best bacterial isolate for controllingyam leaf spot was identifiedby morphologicaland biochemical tests, accordingto Norris et al. [19], Schaad[23], and Sneath [27]. Tests includedGram reaction; cell morphology;endosporeformationand location; growth at 45C, pH 5.7, or 7% NaC1;anaerobic growth in glucosebroth; utilizationof citrate; acid productionfrom arabinose, mannitol, and xylose; Voges Proskauertest and starchhydrolysis. A furtheridentificationwas performedby the Funda~fioTropicalde Pesquisas e Tecnologia"Andr6 Tosello," locatedin Campinas,S~o Paulo State, Brazil.

Results

Laboratory Experiments Among 162 epiphytic bacteria isolated from the yam phylloplane, 39 behaved as C. eragrostidis antagonists by the paired culture test. They were grouped in four frequency intervals (Fig. 1). Only two bacterial isolates displayed mycelial growth inhibition greater than 41%. The IF-24 isolate showed the largest inhibitory effect. The cellophane test showed that all 80 bacterial isolates produced extracellular nonvolatile metabolites that inhibited C. eragrostidis mycelial growth. Among those, only 17 isolates induced more than 75% inhibition (Table 2). The IF-26 metabolite(s) showed the greatest inhibitory effect (93.4%). In the absence of bacterial cells, C. eragrostidis conidia produced long germination tubes after 6 h of incubation in sterile distilled water. Paired suspensions indicated that IF-82 was the greatest antagonist of conidial germination, inhibiting germination by 99.2% (Table 3).

Greenhouse Experiments There was a significant difference in disease severity reduction among the antagonist application periods, evaluated 14 days after the pathogen inoculation (Table 4). Best results were obtained when antagonists were applied 3 days before the phyto-

Antagonism to Curvularia Leaf Spot

105

Table 2.

Mycelial growth inhibition (MGI %) of Curvularia eragrostidis induced by extracellular metabolites produced by epiphytic bacteria from the yam phylloplane Bacterial isolate

MGI (%)a

IF-26 IF-82 IF-88 IF-36 IF-42 IF-76 IF- 164 IF-80 IF-81 IF-40 IF-168 IF-109 IF-85 IF-14 IF-71 IF-24 IF- 124

93.4 a 88.6 ab 88.3 ab 88.3 ab 88.0 ab 86.2 abc 85.0 abc 82.6 abe 82.6 abe 81.4 abc 80.5 abc 80.0 bc 78.4 bcd 77.8 bcd 77.0 bcd 76.7 bcd 76.0 bcd

aAverage of three replications. Means followed by the same letter do not differ (P = 0.05) accordingto Tukey's test pathogen inoculation. When the three antagonist application periods were considered together, the IF-82 isolate showed the greatest control (58.3%). Considering the five best antagonists, application simultaneously with the phytopathogen inoculation produced the best results, except for IF-36. Antagonist application 3 days before C. eragrostidis also provided good results, particularly for IF-88 and IF-36 which showed 70.8% reduction of leaf spot severity. IF-88 showed less efficiency when applied 3 days after the pathogen inoculation. Table 5 shows the persistence of the antagonistic action. When all application periods were considered together, IF-82 had significantly greater persistence (96.3%) than the others. When all five isolates were considered together, the greatest maintenance of antagonistic action was observed in the treatments applied 3 days before and at the time of inoculation, which differed statistically from that of the treatment 3 days after. The bacterial isolates did not show significant differences in the persistence of the antagonistic action in the first two application periods. IF-82 exhibited the greatest persistence with 100% values in the first two application periods. The results obtained from greenhouse and laboratory tests were similar and suggest that IF-82 has great potential in the biocontrol of Curvularia leaf spot of yam.

Identification of the IF-82 Bacterial Isolate The IF-82 isolate was identified as Bacillus subtilis Cohn.

106

S.J. Michereff et al.

Table 3. Conidiagermination inhibition (CGI %) of Curvulariaeragrostidisin suspensions paired with epiphytic bacteria from the yam phylloplane

Bacterial isolate

CGI (%)a

IF-82 IF-88 IF-109 IF-76 IF-80 IF-36 IF-42 IF-164 IF-24 IF-40 IF-26 IF-14 IF-85 IF-71 IF-124 IF-81 IF-168

99.2 a 98.2 a 96.2 ab 90.4 bc 89.0 bc 87.3 bc 83.8 c 55.1 d 42.0 de 35.7 e 33.0 ef 31.3 ef 27.6 ef 19.2 f 4.4 g 0.5 gh 0.0 h

aAverage of four replicates. Data were transformedto arc sin x/-~-d-6. Means followed by the same letter do not differ (P = 0.05) according to Tukey's test

Discussion

The inhibitory effects of bacterial isolates on C. eragrostidis mycelial growth shown in the paired culture method indicate the natural existence of a large number of resident organisms with antagonistic potential on the yam phylloplane. Ability of the antagonist to grow and survive on the leaf surface is an extremely important factor for phylloplane disease biocontrol [2, 4, 7, 9, 10, 13, 17]. The formation of inhibition zones between phytopathogen and antagonist colonies in the paired test indicates the production of bacterial extracellular metabolites that characterizes the antibiosis mechanism discussed by Cook and Baker [12]. However, the antifungal activity observed "in vitro" does not always correspond to disease reduction "in vivo." The "in vitro" production of nonvolatile metabolites has frequently been used as a primary step in the selection of potential biocontrol agents [1]. The cellophane test suggested the antibiosis of the bacterial extracellular, nonvolatile, and diffusible metabolites against C. eragrostidis. The variation among the bacterial isolates in ability to inhibit growth of C. eragrostidis might be due to genetic variability or to environmental factors such as nutrition or pH [2, 14]. The 75% C. eragrostidis mycelial growth inhibition points to a strong fungistatic activity of the bacterial metabolites, as observed for the interactions of B. subtilis versus Bipolaris oryzae (Breda-de-Haan) Shoemaker [16] and Moniliniafructicola (Wint.) Honey [20].

Antagonism to Curvularia Leaf Spot

107

Table 4. Influence of the time of application of epiphytic bacteria on yam leaf spot severity reduction (DSR %). Disease severity was evaluated 14 days after Curvularia eragrostidis inoculation Bacterial

Application period (DSR %)a

isolate

- 3

IF-82 IF-14 IF-88 IF-36 IF-164 IF-42 IF-24 IF-76 IF-109 IF-71 IF-26 IF-124 IF-80 IF-81 IF-85 IF-40 IF-168

66.7 58.3 70.8 70.8 34.5 62.5 62.5 54.2 54.2 33.3 29.2 8.3 33.3 37.5 62.5 16.7 4.2

a abc a a bcde ab ab abcd abcd cdef def fg cdef hcde ab efg g

0 A A A A B A A A A A A B A A A A A

75.0 66.7 75.0 25.0 66.7 33.3 41.7 41.7 25.0 41.7 41.7 58.3 25.0 25.0 0.0 25.0 0.0

+3

a ab a de ab cd bcd bcd de bcd bcd abc de de de e e

A B A B A B B A B A A A AB A B A A

33.3 41.7 4.2 41.7 29.2 20.8 0.0 0.0 0.0 0.0 4.2 0.0 8.3 0.0 0.0 16.7 0.0

ab a cd a abc abcd d d d d cd d bcd d d abcd d

B C B B A B C B C B B B B B B A A

aAverage of three replications. Means followed by the same letter (lower-case letters for rows and upper-case letters for columns) do not differ (P = 0.05) according to Tukey's test. Application periods: - 3 , three days before; 0, at the same time; +3, three days after

Table 5. Persistence of antagonistic action ofepiphytic bacteria from the yam phylloplane applied at different periods, on the Curvularia leaf spot biocontrol Bacterial isolate IF-82 IF-164 IF-14 IF-36 IF-88

Application period (PAA %)a - 3 100.0 77.8 72.2 76.7 82.2

0 aA aA aA aA aA

100.0 75.0 66.7 66.7 94.4

+3 aA aA aA aA aA

88.9 55.6 66.7 50.0 0.0

aA abA abA bA cB

aAverage of three replications. Means followed by the same letter (lower-case letters for rows and upper-case letters for columns) do not differ (P = 0.05) according to Tukey's test. Application periods: - 3 , three days before; 0, at the same time; +3, three days after

The fast germination of C. eragrostidis in sterile distilled water demonstrated its low dependence on exogenous nutrients to form infection structures [17]. The strong inhibitory effect of bacterial isolates, especially IF-82, on C. eragrostidis conidium germination, reinforces their potential for yam leaf spot biocontrol. According to Blakeman [6], spore germination on the phylloplane is a critical stage in pathogen development in which the pathogen is vulnerable to other microrganisms that may influence the success or failure of the infection process. The antago-

108

S.J. Michereffet al.

nistic activity of the bacterial isolates on pathogen conidium germination is probably related to the production and diffusion of extracellular nonvolatile metabolites. This effect is usually associated with antibiosis by lethal action or chemical toxicity [12]. In the present study the application of bacterial isolates at the same time as pathogen inoculation or before pathogen inoculation presented better efficiency of yam leaf spot control under greenhouse conditions, as compared to applications after the pathogen. Similar results were reported for B. subtilis controlling Uromyces phaseoli (Reben.) Wint. in beans (Phasealus vulgaris L.) [3], as well as for Cylindracladium scaparium Morgan in eucalyptus leaves (Eucalyptus grandis Hill ex. Maid, and E. urophyla S.T. Blake) [5]. The efficiency of preventive bacterial application is probably associated with C. eragrostidis behavior during the prepenetration phase on the host surface, as demonstrated by Blakemam [7] for other pathogens. Considering only the most efficient antagonists, the best yam leaf spot control was achieved by same-time application. Similar data were obtained for B. subtilis controlling Alternaria alternata (Fr.) Keissl. in tobacco (Nicatiana tabacum L.) [15]. Considering the levels of biocontrol of other phylloplane diseases achieved with epiphytic bacteria [16, 24, 28], the 75% disease severity reduction given by IF-82 and IF-88 isolates applied at the same time as the pathogen is an excellent result. The IF-88 isolate showed very little efficiency when applied after the pathogen inoculation, indicating its low colonization and competitive abilities following pathogen establishment, and/or a low vulnerability of the pathogen once established. In the present study, the protective and curative actions of the bacterial isolates seem to be associated with the release of inhibitory substances on the phylloplane Antibiosis caused by epiphytic bacteria on spore germination of pathogenic fungi is considered significant in disease biocontrol [4, 8, 16, 24, 28]. The persistence of antagonistic action is an important factor in phylloplane disease biocontrol [10, 13]. The high level of persistence of antagonism of IF-82 suggests its strong ability to colonize the phylloplane and to form a biologically active population whose metabolites remain active in concentrations sufficient to inhibit the pathogen. Some researchers [8, 11, 21], however, have described the instability of antibiotics and metabolites produced by phylloplane microorganisms, caused by adsorption, immobilization, oxidation, dilution, and repeated condensation and water evaporation. IF-82 may be considered as a potential yam leaf spot control agent, confirming the hypothesis that the isolation of antagonists for phylloplane disease biocontrol may be accomplished from the natural host microflora [2]. The identification of the IF-82 isolate as B. subtilis is significant since the antifungal activity of B. subtilis metabolites has been reported [10, 20, 25]. The ability of B. subtilis to form endospores facilitates its survival and metabolic activity under adverse conditions such as variation in the phylloplane environment [17].

Acknowledgment. The authorsare gratefulto Dr. DirceYanofromthe Funda~toTropicalde Pesquisas e Tecnologia"Andr6Tosello,"Brazil, for the identificationof isolateIF-82.

Antagonism to Curvularia Leaf Spot

109

References 1. Andrews JH (1985) Strategies for selecting antagonistic microorganisms from the phylloplane. In: Windels CE, Lindow SE (eds) Biological control on the phylloplane. The American Phytopathological Society, St. Paul, pp 31-44 2. Andrews JH (1992) Biological control in the phyllosphere. Ann Rev Phytopathol 30:603-635 3. Baker CJ, Stavely JR, Thomas CA, Sasser M, Macfall JS (1983) Inhibitory effect of Bacillus subtilis on Uromycesphaseoli and on development of rust pustules on bean leaves. Phytopathology 73:1148-I 152 4. Bettiol W (1991) Controle biol6gico de doen~as do filoplano. In: Bettiol W (org) Controle biol6gico de doen~as de plantas. Embrapa-Cnpda, Jaguarifina, pp 35-56 5. Bettiol W, Auer CG, Carnargo LEA, Kimati H (1988) Controle da mancha foliar de Eucaliptus grandis e E. urophyUa induzida por Cylindrocladium scoparium corn Bacillus sp. Sum Phytopathol 14:210-218 6. Blakeman JP (1982) Phylloplane interactions. In: Mount MS, Lacy GH (eds) Phytopathogenic prokaryotes, vol 2. Academic Press, New York, pp 308-333 7. Blakeman JP (1985) Ecological succession of the leaf surface microorganisms in relation to biological control. In: Windels CE, Lindow SE (eds) Biological control on the phylloplane. The American Phytopathological Society, St. Paul, pp 6-30 8. Blakeman JP, Brodie IDS (1976) Inhibition of pathogens by epiphytic bacteria on aerial plant surfaces. In: Dickinson CH, Preece TF (eds) Microbiology of aerial plant surfaces. Academic Press, London, pp 529-557 9. Blakeman JP, Fokkema NJ (1982) Potential for biological control of plant diseases on the phylloplane. Annu Rev Phytopatho120:167-192 10. Blakeman JP, Brown AE, Mercer PC (1992) Biological control of plant diseases--present and future trends. Pesq Agropec Bras 27:151-164 11. Boudreau MA, Andrews JH (1987) Factors influencing antagonism of Chaetomium globosum to Venturia inaequalis: a case study in failed biocontrol. Phytopathology 77:1470-1475 12. Cook RJ, Baker KF (1983) Nature and practice of biological control of plant pathogens. The American Phytopathological Society, St. Paul 13. Elad Y (1990) Reasons for the delay in development of biological control of foliar pathogens. Phytoparasitica 18:99-105 14. Fravel DR (1988) Role of antibiosis in the biocontrol of plant diseases. Annu Rev Phytopathol 26:75-91 15. Fravel DR, Spurr Jr HW (1977) Biocontrol of tobacco brown spot disease by Bacillus subtilis subsp, mycoides in a controlled environment. Phytopathology 67:930-932 16. Handa HP, Gangopadhyay S (1983) Control of rice helminthosporiose with Bacillus subtilis antagonistic towards Bipolaris oryzae. Int J Trop Plant Dis 1:25-30 17. Knudsen GR, Spun" Jr HW (1988) Management of bacterial populations for foliar disease biocontrol. In: Mukerji KG, Garg KL (eds) Biocontrol of plant diseases, vol 1. CRC Press, Boca Raton, pp 83-92 18. Nigam N, Mukerji KG (1988) Biological control---concepts and practices. In: Mukerji KG, Garg KL (eds) Biocontrol of plant diseases, vol 1. CRC Press, Boca Raton, pp 1-13 19. Norris JR, Berkeley RCW, Logan NA, O'Donell AG (1981) The genera Bacillus and Sporolactobacillus. In: Stan MP (ed) The prokaryotes. Springer-Verlag, Berlin, pp 1711-42 20. Pusey PL (1989) Use of Bacillus subtilis and related organisms as biofungicides. Pest Sci 27:133140 21. Rai B, Singh DB (1980) Antagonistic activity of some leaf surface microfungi against Alternaria brassicae and Dreschslera graminea. Trans Br Mycol Soc 75:363-369 22. Robbs CF (1991) Bactrrias como agentes de controle biolrgico de fitopatrgenos. In: Bettiol W (org) Controle biolrgico de doen~as de plantas. Embrapa-Cnpda, Jaguaritina, pp 121-134 23. Schaad NW (ed) (1988) Laboratory guide for identification of plant pathogenic bacteria, 2nd ed. The American Phytopathological Society, St. Paul 24. Sharma JK, Sankaran KV (1988) Biological control of rust and leaf spot diseases. In: Mukerji KG, Garg KL (eds) Biocontrol of plant diseases, vol 2. CRC Press, Boca Raton, pp 1-23

110

S.J. Michereff et al.

25. Sinclair JB (1989) Bacillus subtilis as a biocontrol agent for plant diseases. In: Agrihotri VP, Singh N, Chaub HS, Singh US, Dwivedi TS (ed) Perspectives in plant pathology. Today & Tomorrow's, New Delhi, pp 367-374 26. Sistema de produ~o para car~ da costa: Agreste Setentrional, Agreste Meridional e Mata Norte. EMATER/IPA, Recife, Brazil 27. Sneath PHA (1986) Endospore~forming Gram-positive rods and cocci. In: Sneath PHA, Mair NS, Sharpe ME, Holt JG (eds) Bergey's manual of systematic bacteriology, vol 2. Williams & Wilkins, Baltimore, pp 1104-1207 28. Spun" Jr HW, Knudsen GR (1985) Biological control of leaf diseases with bacteria. In: Windels CE, Lindow SE (eds) Biological control on the phylloplane. The American Phytopathological Society, St. Paul, pp 45~62 29. Vidaver AK (1982) Biological control of plant pathogens with prokaryotes. In: Mount MS, Lacy GH (eds) Phytopathogenic prokaryotes, vol 1. Academic Press, New York, pp 387-397