'zh I I . IL“ . . . 1.. ... .. v....JVa.N.rJ! .w. . L.” n .uvlm up". . . . -& w t: .nu . 1.. s I 01.1.... . .. 7.. .\. JELykw‘Tu : \C. .x. . u o o g ‘I fir ot’nfl v ‘0‘ c oil-.5 .slhll Pa’lt. luv: 2......“ .. :r .f.t»».h.u«x$flhwnmwwhg . 1K3 ($.13 v¢«.M. 51%-. . ... v.51“. .QInNVIiIFH.“ ,. .. 9411.093”? 31...... ..x:.. .. n . .1. i .. y :1 .. a: Inn”... . . . ...u . ‘nw.m.r..mhm.\fi$.m«nwtdlhb .. .s ‘1. 0 RD“! §g\ h mMfiMMMmW-Mvflfimmyn. ~ 2.4" , ., i. ... n1...mfi. J. . in”? .8. . . 1:13 . p .‘C\ t; ‘1‘ . a. . :3 s: Q‘ 1 t. .: oczl... Lu....o: . - Vou¢ac 0) I5:- .. . 35.93%. h {3:} . 1! ii: J' I“ 1 ‘i 1.. .' ' ‘3 0 .. .H wuhnvflw HMWJIH 2. 31¢. WHWWJ .M.m ON 0 “Jam: 5. HM M...n.u%.~.vl....hh .%m§fi..wav. V: o, 55.....- INHVNfiJ.Ho ctH‘h‘ . K .15.“- . . ... v.5. «ahdfiuuua :8. . NURHFF! : . .. "6.7!. .....\ .nufiwnh .. . . vffl. 11.1.... v .1 . . I . n .4. u . . N . C \. 34-11:,» . . . p.” . . n“. . . . . s 4.. t u: v%.. ¢r90 . .... . . .... , . . . . “WNW-Ea. ‘ E fiydvAAgw fidowfiri I . . . . . . . . . . n 7 an” (chi: h. . H. . .. . . 3. p... . . .v I . . . . In C A . V . ... a... T . . 1 U; . b7 H: .ncnmgmhaxwm» «MW. ... H... . . . Lémfiflflkkmmf mm. @114. .. van. 4!. i .3 u v r o. --.\.‘..o.‘ w! {3.3} Ito-.I 00". . .. 00 T . 1" lo! II Q‘m.‘ .N 2’: It‘s.” ,oOOV‘LIOW. 0‘ ,Va ¢ % . "c.1161 .fi’vu\ocns-.Q._. o hoahovu.:|¢ lufi I VQI.’ is u . .. Ru. -4 A .U? 1 .. .4 I! .. ,, ,, :24wnumwzfi. “#me IN“. N “mutt. ’Wjipdtfl ”Muttssh.‘ 412 ....m.nv.mmm....ou.la . . 4.. immnflnrlfiti $5110.: flnwunuhh. it... « «1 . v‘YIIr v. V). .45301- . . $.34 . It a 6. r .Io..\ . '10.... 4 1:. I: o v I." canhlflfllnohl. '\ .. . | '5 .. til. at. r. 1. u 4‘ ‘ II I b I . I Ola..tl.l‘o . ..vo . 1‘ .. t I. . V f... . 313v. . L...YIL..r....d. . ‘9... [IVY-.00 \I 5.60:“. '. ,.!Ca'ol. I»! IAIIfiVXI . .lll .\.Ll .- .Sscol .v. :7.“ . _ I. (Int. ‘14 v. . 1'74". 0U. I... . “walla. fling. I ...1\P.\¢h«:v ’1 3 viii...) v - f0..-Ai..‘i. ! a J .0 3.. It‘x$¢ll§. \n '90.). (O?! I v. ‘31..-" I'Ak‘: 3" g 10" A. . :rlflor .‘svll'lo v fin~¢clh¢rllu . :59! (It: ill! NFL... .25.} as. .r.:‘ .3! rut. ll n Novvlnt...t Ika-u. oil)! 0% +5‘.$nu “WA. Tl? .- 5.1!. Pd 1 '1 'I A .Ié‘l ‘A O I r It: I. hnm‘mx. irlr. , . .lx L oozus. -.v.t..ml .N... . t... . 3.14an 3%,... “than. 14h. r-.. I ': vtllu . . (4“. ‘1'!- .6t' I t, ‘l [a bi .txvt v I}. W.lo Fun}... 1.3. {Iain rt. ....L...... l1 f.‘l-i’ot 1! III- 515173.5QO9IWB! r“. .‘ fillgt. I4 grog-‘2 l‘rv‘ l!\q|ct~!\v I‘o’i’ “VI-v. to VAL". . . ‘ . I. )‘V. a. {othWLLloiJll.t§.lf. . fintfilyim “Mod? ..4. . . . . . , .w.. . I 'om A . n ~ Ill. 0‘ Al | l 3 u ' {a p.” . a. . .t. v‘l’lltufiflnvlkt V 3. ~ '4“ .IO:A\D1I§|\O‘ fig. .13. ..finnwauunwin..§.u....n..:....... - . . . . $32.--. .fikfiw _ mu. . . . .8153... 1-13.... %§.§:l . I. . ..<.............. . , n 1.13... .9...» .5: . v.;....(..f...-....1.fi.,.. 2.. .r: .9 E! I :.t , m . S? t L 3 .- . .4- 1&3. .- ....c..... L . 11.5.)Ivo‘... . J. $7.: ..v. ....l.. ..u.. .. V... .. . .~. ...a... «a Z. . » ... x . . .1. uni... .. ‘ . . . . A . . : n . . . v a V». . . c ..u . . I . Dc I . ‘ 1.43.1.1... . x. .. ..v . . r U o .. . .QII~ OI . . I! {in IIHIUHIIillIllllllll‘llllllllllllllWNWllllllllll’ll 31293 01591 3746 LIBRARY Michigan State University This is to certify that the thesis entitled Development of an improved semi-selective medium for Xanthomonas campestris pv. phaseoli and its use in characterizing resistant bean germplasm. presented by Robert Bandoma Mabagala has been accepted towards fulfillment of the requirements for M.S. degree in Plant Pathology (mm \oxwmw major professor DateWT 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU ‘ LIBRARIES .—:l__ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. DEVELOPMENT OF AN IMPROVED SEMI-SELECTIVE MEDIUM FOR XANTHOMONAS CAMPESTRIS PV. PHASEOLI AND ITS USE IN CHARACTERIZING RESISTANT BEAN GERMPLASM BY Robert Bandoma Mabagala A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Botany and Plant Pathology 1987 ABSTRACT DEVELOPMENT OF AN IMPROVED SEMI-SELECTIVE MEDIUM FOR XANTHOMONAS CAMPESTRIS PV. PHASEOLI AND ITS USE IN CHARACTERIZING RESISTANT BEAN GERMPLASM by Robert Bandoma Mabagala An improved solid semi-selective medium (M-SSM) for isolation of Xanthomonas campestris pv. phaseoli (XCP) was developed by modifying the liquid semi-selective medium (SSM) to contain starch, dyes, and agar. M-SSM was equal or superior to the other semi—selective media (MXP and DSX) for isolation of XCP from various infected bean material and from infested soil. M—SSM was used to monitor population trends of XCP on and in reproductive tissues and levels of seed infection in seven bean genotypes. XCP was recovered from flower buds, blossoms, pods, and seeds of both resistant and susceptible bean genotypes. Epiphytic bacterial populations from reproductive tissues of various bean genotypes were screened for antagonism to XCP. Three isolates exhibited antagonism to XCP in yitrg. in 3139, all three delayed symptom appearance for 2-3 days when spray inoculated prior to XCP. The rate of disease development was significantly reduced. In memory of my parents and sisters. ii ACKNOWLEDGEMENTS I would like to express my sincere appreciation to Dr. Alfred w. Saettler for his constant guidance throughout the course of this study. The friendly atmosphere he created towardéme and his great interest in my work during the period of this study was enjoyable and rewarding. I credit him for his professional talent. I also thank the other members of my graduate committee Drs. Christine Stephens and Jim Kelly for their helpful suggestions during preparation of this manuscript. Moral support given by graduate students and technicians in the laboratory during this study is highly appreciated. The USAID/Title XII Bean CRSP Program which provided support to the author during the course of this work is gratefully acknowledged. I extend my special appreciation to Dr. M.J. Silbernagel, US Principal Investigator for Washington State University/Tanzania Bean CRSP. iii TABLE OF CONTENTS Page LIST OF TABLES ....................................... vii LIST OF FIGURES ......... O 0000000000 00...... 0000000000 x GENERAL INTRODUCTION AND LITERATURE REVIEW ........... 1 CHAPTER I DEVELOPMENT OF AN IMPROVED SEMI-SELECTIVE MEDIUM (M-SSM) FOR ISOLATION OF XANTHOMONAS CAMPESTRIS PV. PHASEOLI MATERIALS AND METHODS ............................... 12 Bacterial strains .............................. 12 Pathogenicity tests........... ................. 14 Test media.................... ............ ..... 14 Development of M-SSM medium...... ........... ... 17 Pure culture studies.... ..... .................. 18 Recovery from naturally infected bean leaves... 19 Recovery from artificially inoculated bean tissue ............................. .......... 20 Recovery of XCP from blossoms of artificially inoculated bean plants....................... 21 Recovery of XCP from dry bean tissue powder... 21 Recovery of XCP from bean dust collected from seedO. ...... .... ....... ......OOOOOOOOOOOOOOOO 22 Effect of media age on recovery of XCP......... 22 Recovery from artificially infested soil ....... 23 RESULTS OOOOOOOOOOOOOOOOOOOOOOOOOO O OOOOOOOOOOOOOOOOOO 25 Pathogenicity test of XCP isolate .............. 25 Media evaluated .............. . ................. 25 Time required for colonies to develop.. ........ 28 Development of improved semi-selective medium (M- SSM).... ..... .. ......... . .......... 3O Plating efficiency.... ...... . ........ .... ..... 31 Recovery from naturally infected bean leaves. 37 Recovery from artificially inoculated bean tissue ...... O 00000000 000...... ....... ... ..... 37 Recovery from blossoms of artificially inoculated bean p1ants...... ................. 40 iv Recovery of XCP from dry tissue samples.. ...... Recovery of XCP from bean dust collected from seed ...... 0000...... ..... .... OOOOOOOOOOOOOOOO Effect of media age on recovery ................ Recovery from artificially infested soil ....... DISCUSSION ................ ...... .................... CHAPTER II POPULATION TRENDS OF XANTHOMONAS CAMPESTRIS PV. PHASEOLI IN REPRODUCTIVE TISSUES OF DIFFERENT BEAN GENOTYPES IN THE FIELD. INTRODUCTION ........ . ............................... MATERIALS AND METHODS.... ..... .. .................... Location of the experiment ..................... Bean genotypes............................ ..... Experimental design..................... ....... Bacterial isolates and inoculum preparation.... Inoculation of plants ................... ....... Field evaluation of disease reactions .......... Measurement of bacterial population ..... . ...... Estimation of epiphytic bacterial contaminants associated with bean reproductive tissues.... Pathogenicity tests........ ............... ..... Assessment of seed infection ................... Statistical analyses ........................... RESULTS ..... ' ........................................ Disease reactions .............................. Bacterial populations ..... ... .................. Epiphytic bacterial contaminants............... Pathogenicity tests ....... ..... ................ Seed infection.... ............................. DISCUSSION .......................................... CHAPTER III ISOLATION OF EPIPHYTIC BACTERIAL ANTAGONISTS FROM VARIOUS BEAN GENOTYPES AND THEIR POTENTIAL FOR CONTROL OF XANTHOMONAS CAMPESTRIS PV. PHASEOLI INTRODUCTION... ..................................... MATERIALS AND METHODS ............................... 4O 43 43 46 SO 89 92 Bacterial isolates .......... .. ................. 92 Pathogenicity tests................ ............ 93 In vitro screening for antagonism .............. 93 In vivo screening for antagonism ............... 95 Statistical analysis............. .............. 96 RESULTS ............................................. 97 Pathogenicity tests................ ............ 97 In vitro screening for antagonism .............. 97 In vivo screening for antagonism ............... 102 DISCUSSION .......................................... 106 LITERATURE CITED .................................... 109 vi Table 10 LIST OF TABLES Identity of bacterial isolates used for evaluation of test media..... ..... .. ........ Pathogenicity tests of Xanthomonas campestris pv. phaseoli (Xp) and fuscans variant (pr) isolates used for evaluation of selective media ............... . .......... Recovery of Xanthomonas campestris pv. phaseoli from naturally infected bean leaves on MXP, SSM, DSX’ and YCAOOOOOOOOOOOOOOOOOOO Plating efficiency of various isolates of Xanthomonas campestris pv. phaseoli (Xp) and fuscans variant (pr) on MXP, SSM, and M_SSM ooooooooooo . ....... 0....0........ ...... Comparative colony development of Xanthomonas campestris pv. phaseoli (Xp) and fuscans variant (pr) on four test media five days after planting......................... Recovery of various bacterial isolates on MXP, SSM, M-SSM andYCA......00..00.....0... Recovery on three semi-selective media of Xanthomonas campestris pv. phaseoli from naturally infected bean leaves.............. Recovery on four test media of various isolates of Xanthomonas campestris pv. phaseoli from artificially inoculated bean leaveSOO ........ 0..........OOOOOOOOOOOOOOOO. Recovery of Xanthomonas campestris pv. phaseoli from flowers of artificially inoculated bean plants grown in the green house ....................................... Recovery on test media of Xanthomonas campestris pv. phaseoli from infected dry bean tissue stored for nine months at room temperature ............... ........ .......... vii Page 13 26 27 32 34 35 38 39 41 42 Table Page 11 Recovery of Xanthomonas campestris pv. phaseoli from bean dust collected from seed grown in Puerto Rico.......... ........ 44 12 Effect of media age on percent recovery of Xanthomonas campestris pv. phaseoli ........ 47 13 Recovery on four media of various isolates of Xanthomonas campestris pv. phaseoli and other bacteria from artificially infested soil ........ .. ..... ............ ............ 48 CHAPTER II 1 Leaf reactions of seven bean genotypes to Xanthomonas campestris pv. phaseoli (MI-17) under field conditions. 0 O O O O O O O O O O O O O 0 O O O O O 68 2 Reactions of pods on plants of seven bean genotypes at physiological maturity to Xanthomonas campestris pv. phaseoli (MI-l7) under field conditions..................... 70 3 Surface populations of Xanthomonas campestris pv. phaseoli (MI-l7) on flower buds and open flowers of various bean genotypes grown in the field ..... . ......... 72 4 Internal populations of Xanthomonas campestris pv. phaseoli (MI-l7) in flower buds and open flowers of various bean genotypes grown in the field ............... 73 5 Surface populations of Xanthomonas campestris pv. phaseoli on flat and bumpy pods of various bean genotypes grown in the field ............... . ....... ........... 75 6 Comparative population levels of bacterial contaminants from flower buds and open flower washates of various bean genotypes grown in the field.. ..... .................. 77 7 Population levels of bacterial contaminants in washates from flat and bumpy pods of various bean genotypes grown in the field.. 78 viii Table Percent seed infection of various bean genotypes inoculated with Xanthomonas campestris pv. phaseoli under field conditions.......... ........ ........ ....... CHAPTER III Epiphytic bacterial contaminants isolated from bean plants grown in the field........ In vitro screening of epiphytic bacterial contaminants from bean plants for ability to inhibit growth of Xanthomonas campestris pv. 213389011(MI‘17)......................o In vivo screening of epiphytic bacterial antagonists for ability to control Xanthomonas campestris pv. phaseoli in beanSOO ..... .........OOOOOOOOOOOOOOOO ...... In vivo screening of epiphytic bacterial antagonists for ability to control Xanthomonas campestris pv. phaseoli in beans ........ ...................... ..... ... ix Page 81 98 101 103 104 Figure LIST OF FIGURES Page CHAPTER I Comparative growth and colony characteris- tics of Xanthomonas campestris pv. phaseoli on SSM, MXP, M-SSM, and YCA (control), after five days incubation at 27 i 1 C.... 29 Starch hydrolysis zones around colonies of Xanthomonas campestris pv. phaseoli on M—SSM after four days of incubation at 27 i l C ..... ...... ........ .......... ..... 36 Recovery of Xanthomonas campestris pv. phaseoli on YCA (control), SSM, MXP, and M-SSM from bean dust collected from seed grown in Puerto Rico. Photographs taken after 5 days incubation at 27 i l C...... 45 CHAPTER II Recovery of Xanthomonas campestris pv. nhaseoli from bean seed harvested from plants of line I-84100 grown in the field and spray inoculated when 18 days old. Medium was M-SSM and plates were incubated 3 days at 27 i 1 C.... ...... .............. 80 CHAPTER III Pathogenicity test of epiphytic bacterial contaminants isolated from reproductive tissues of beans grown in the field. (A) Pathogenic XCP (MI-l7). (B) Bacterial contaminant No. 22.... ...... .............. 99 Zones of Xanthomonas campestris pv. nhaseoli inhibitions around colonies of bacterial antagonist No. 8. Bacterial contaminants No. 1 in the same lawn is without inhibition zones. NBGA was used for both basal layer and the overlay medium (containing 1% agar).... ........... 99 INTRODUCTION AND LITERATURE REVIEW Xanthomonas campestris pv. phaseoli (E.F. Smith) Dawson and I; campestris pv. phaseoli var. fuscans (Burkh) Starr and Burkh, causal agents of bean common and fuscous blight, respectively, continue to be serious problems in many bean growing areas (9, 15, 79). Both bacteria are now recognized as I; campestris pv. phaseoli (Smith) Dye (2) and hereafter will be referred to collectively as XCP. These bacterial pathogens frequently occur together in a field and probably on the same plant, increasing the difficulty of associating yield losses with a specific pathogen (82). The world-wide distribution of these pathogens is in part associated with the ability to infect seed. Internally infected seed has been reported to be the main source of primary inoculum for common and fuscous blight bacteria (9, 61, 63, 80, 83, 84). Presence of the bacteria in seed is an important means of survival and dissemination in time and space. Studies indicate that XCP can survive as long as 35 years in infected bean seed under laboratory conditions (56). In the field, a very low level of seed infection can give rise to high disease incidence. Epidemiological studies indicate that one infected seed in 10,000 is sufficient to result in an epiphytotic of blight in the bean crop (72, 78). Much research has been directed toward maintaining bean seed stocks free of XCP contamination and developing methods for detection of seed-borne infection. Control of XCP is based on seed certification programs to maintain clean seed stocks (16, 75). There is no doubt that such programs have been successful in reducing seed infection by XCP, nevertheless, disease outbreaks persist and in Michigan, some fields are rejected annually for certification (63). In order to design effective disease control measures, it is often necessary to know where plant pathogenic bacteria survive between growing seasons; as part of their life cycle. To detect survival of plant pathogenic bacteria, adequate techniques are needed, which must be able to detect low population densities of these bacterial pathogens under natural conditions (23). Various techniques have been developed for routine detection of bacterial blight infection in bean seeds. Several workers have shown that phages active against Xanthomonas campestris pathovars have a high degree of specificity and can be used as a detection tool for many seed-borne bacterial pathogens (38, 39). Sutton and Wallen (52, 71) studied 310 isolates of XCP and identified eight phage types. There was some suggestion of a relationship between phage type and geographic origin of isolates, and also virulence. Highly specific phages have been used to detect plant pathogenic bacteria in material from which sampling by conventional means would often give false negative results. The presence of the pathogen is presumed when the titre of the specific phage increases after incubation with test sample for several hours (52). The technique was first used by Katznelson (28) and later by Katznelson and Sutton (28, 29, 30, 52). However, bacteriophage techniques have not been widely adopted by other workers. While phage is a simple and rapid method for detection and identification of plant pathogenic bacteria, extensive tests must be run to determine specificity. According to Schaad (58) a major disadvantage of the direct phage—plaque method for detecting and identifying bacteria in seed is the lowered sensitivity when large numbers of other bacteria are present. He added that, perhaps the greatest disadvantages of using phage are the unusual lack of species-specific host range and the resistance of some bacterial strains. False negative results have been observed when using the phage technique with Pseudomonas syringae pv. phaseolicola on bean seed. These inconsistencies were thought to be due to a deficiency in the bacterium-phage interaction, or to the presence of a phage inhibitor in seed macerates (52). Katznelson (28) also observed that size of the sample and degree of seed infection were crucial for successful tests. The false negative results of Katznelson's phage plaque method resulted in the epiphytotic of fuscous blight in Canada in 1962. The pathogen presence in increasing levels in seed field escaped detection for a number of years because the specific bacteriophages used in the test were unable to detect this strain of the pathogen (66, 78). The use of serology to detect and identify seed-borne bacteria directly in enrichment culture of bean seeds and other infected plant material has been reported (1, 43, 57, 58, 74, 75). The main advantages of serological tests are speed and low cost. Schaad (58) reported that the serological test most suited for identification of seed- borne bacteria is immunofluorescence. As little as 10 microliters of antiserum is needed for each sample and results are available in a single day. However, a major disadvantage of serology is that the method for each organism must be thoroughly studied, and the specificity and reliability of any serological method depend on the specificity of the antiserum (58). Grow-out tests have been used more than any other method to detect the presence of common and fuscous blight bacteria in bean seeds. Seeds are simply sown in the field or in the greenhouse and the development of disease recorded (58). However, field plot or greenhouse examination of seed lots, while offering a true picture of performance of individual seed lots, does not provide results when they are most needed prior to planting (58, 66, 79). Moreover, grow- out tests are not suited for testing numerous commercial seed lots because of the long time and considerable space required. Such tests are also complicated by other seed- borne bacteria and fungi (58) and by weather conditions which influence development of disease symptoms. Another promising technique for detection and identification of seed-borne bacterial pathogens is the use of DNA probe hybridization (32, 47, 49) which is one of the powerful tools of modern biotechnology. As a result of developments in recombinant DNA technology, there now exists a means of searching for specific nucleotide sequences in a mixture of DNA molecules. The technology was made possible by the discovery of restriction endonucleases, enzymes capable of recognizing specific DNA sequences and making specific cuts in the DNA at these sequences (47). The fragments of DNA generated by these enzymes can be enzymatically spliced into bacterial plasmids that accept foreign pieces of DNA and can be propagated in bacteria, like Escherichia coli. The recombinant plasmids can be produced in large amounts by the bacterial host. Restriction enzymes are again used to release the cloned DNA fragments from the plasmid DNA. Such DNA fragments are then used as DNA probes to seek out and hybridize to complementary sequences in infected samples (32, 47). Although this technique was developed for medical sciences, it is gaining recognition in phytopathological studies and may soon be available for use in a number of laboratories. While the real and potential advantages are obvious, DNA probe hybridization has several disadvantages. Preparing the nucleic acids for hybridization is tedious and lengthy, and the hybridization assay procedure is time consuming (32). Furthermore, in order to be useful as probes, DNA-fragments must be labelled in some way. The most commonly used markers are nucleotides tagged with radioactive isotopes. Radioactivity permits very sensitive detection, but there are serious disadvantages related to cost, radiation hazard, proper disposal, and the short half- life of the most useful isotopes (47). Despite all these advances in diagnostic and detection techniques, there is no doubt that many of the traditional detection and diagnostic methods will remain important methods for bacterial detection (4). Several problems are associated with isolation and detection of phytopathogenic bacteria using conventional artificial media. The pathogen being sought may grow poorly or not at all. A variety of other bacteria culturally similar to the bacterium of interest may grow faster and interact unfavorably with plant pathogens in the media through competition for nutrients, production of adverse pH or creation of an antibiotic effect. In addition, low concentration of the bacterial pathogen associated with non-pathogens in diseased tissue may preclude exenic isolation of the pathogen (13, 19, 24). Cultural problems such as these can be overcome through the use of selective media, which exclude unwanted saprophytes while permitting the growth of the pathogen. Selective and semi-selective media are promising tools for detection and identification of numerous seed-borne bacterial pathogens. Many researchers (13, 14, 19, 58) have stressed the potential of selective/semi-selective media in diagnosis because identification on agar is inexpensive, relatively fast and easy to perform. And perhaps more important, selective media result in culture of the suspected pathogen (26, 58). Semi-selective media have been useful for many quantitative population studies on the survival of phytopathogenic bacteria in the soil, in the rhizosphere or on the phylloplane, on seed or in decaying plant material. Media have also been important in understanding of cycles of infection and in recognizing the existence of alternate inoculum sources in the field (19, 24, 41). Various approaches are used to develop and utilize selective and semi—selective media. Selectivity takes advantage of the different biological systems within species and can involve the use of specific carbon and nitrogen sources to stimulate the growth of one species over that of another. Selectivity also results when bacterial enzymes react with constituents in the medium to promote a cultural distinction between two otherwise similar appearing species (18, 19, 21, 44, 70). Addition of antibiotics that inhibit background microflora, and changes in growth temperatures and oxygen requirements are also useful approaches in the development of selective media (7, 8). The relative availability of pure antibiotics during the past decades have permitted studies on specific bacteriostatic and bacteriocidal effects on phytopathogenic bacteria (7, 8, 75). Results obtained using different semi-selective media vary considerably due to differences in methods of testing, different antibiotics used, different conditions of growth and evaluation; and the use of different test organisms (75). When developing a selective or semi—selective medium, one should consider the type of soil or plant tissues that will be used in the isolation procedure. Samples of soil, rhizosphere soil, leaf tissues, and blossoms all support various types of microbial populations that can interfere with recovery of specific phytopathogenic bacteria (14). Growth rates and temperature requirements of the target bacterium will also affect recovery (8). Other factors including cost, availability of ingredients, ease of preparation, storability and reliability of the medium are important to its practical utilization. This is particularly true when media are to be used in laboratories on large scale for diagnostic or quarantine purposes (8, 14, 19). There has been a significant interest in the use of selective and semi-selective media in the recent years as evidenced by a number of papers published on the subject. Many semi—selective media have been developed and used in several laboratories for studying many phytopathogenic bacteria (17, 46, 50) including Xanthomonas campestris pathovars (2, 10, 11, 12, 14, 18, 26, 44, 54, 59, 60, 76). However, few of these media are suitable for studying XCP (75, 76). Despite these intensified research efforts, no single investigator yet appears to have made a comparative study of reported semi-selective media available for isolation and detection of XCP from various sources. Thus the present work was an effort to: 1) Evaluate the performance of reported semi-selective media for recovery of XCP from various sources and to make improvements, if necessary, to increase selective efficiency. Evaluation was based on recovery from a diverse number of materials. 2) Use an improved semi-selective medium (M—SSM) to study population trends of XCP in reproductive tissues of different bean genotypes grown in the field. 1O 3) To isolate epiphytic bacteria from various bean genotypes and determine their potential for control of XCP. CHAPTER ONE DEVELOPMENT OF AN IMPROVED SEMI-SELECTIVE MEDIUM (M-SSM) FOR ISOLATION OF XANTHOMONAS CAMPESTRIS PV. PHASEOLI 11 MATERIALS AND METHODS Bacterial strains. Identity and sources of bacterial strains used in this study are listed in Table 1. Most isolates of XCP were isolated from bean seed supplied by the Michigan Department of Agriculture and from infected bean leaves collected from different bean growing areas within Michigan. Two strains, Sc-4A and LB-2, were kindly supplied by Dr. A.K. Vidaver, Department of Plant Pathology, University of Nebraska, Lincoln, NE. Pseudomonas syringae pv. phaseolicola strains Pp-P23 and P-P8 were isolated from infected bean leaves from Kenya and Tanzania, respectively, supplied by Dr. J. Kelly, Department of Crop and Soil Science, Michigan State University, East Lansing, MI. Other cultures were obtained from stocks maintained in our laboratory. Working stock cultures were maintained in 0.01 M phosphate buffer/glycerol (40%) mixture stored at 5-7C; others were preserved by routine lyophilization using the procedures described by Sly (68). To recover cultures, several drops of the suspension were aseptically transferred to a fresh YCA agar plate and incubated at 27 1 1C. Lyophilized cultures were recovered by adding several drops 12 Table 1. Identity of bacterial isolates used for evaluation of test media Bacterium identity Lab Isolated from Source where Identification tissue (cultivar) obtained # Xanthomonas campestris pv. phaseoli (fuscans variant) -MI-1 Seed (c-20) MDA(53714)a 5.9 pv. phaseoli MI-2 Seed (Neptune) MDA(52300) 5.9 pv. phaseoli MI-3 Seed (?) Hunger. MI 5.9 pv. phaseoli MI-4 Seed (Seafarer) MDA(53938) 5.9 pv. phaseoli MI-S Seed (Fleetwood) MDA(53402) 5.9 pv. phaseoli(fuscans variant)MI-6 Seed (Seafarer) Botany farm('85) 5.9 pv. phaseoli MI-7 Seed (Seafarer) MDA(12408) 5.9 pv. phaseoli MI-8 Seed (Seafarer) MDA(53419) 5.9 pv. phaseoli MI-9 Seed (Tuscola) MDA(53061) 5.9 pv. phaseoli MI-lO Seed (Neptune) MDA(51853) 5.9 pv. phaseoli MI-ll Seed (C-20) MDA(53724) 5.9 pv. phaseoli MI-12 Seed (Neptune) MDA(51859) 5.9 pv. phaseoli MI-13 Seed (Navy) MDA(53481) 5.9 pv. phaseoli MI-14 Seed (?) MDA(61533) 5.9 pv. phaseoli(fuscans variant)MI-15 Seed (Sanilac) MDA(61807) 5.9 pv. phaseoli(fuscans variant) MI-16 Seed (fleetwood) MDA(61806) 5.9 pv. phaseoli MI-l? Leaves (C-20) Bay City. MI 5.9 pv. phaseoli Sc-4A ? Scottsbluff.NEb 5.9 pv. phaseoli LB 2 ? Nebraskab Pseudomonas syringae pv Phaseolicola Pp-MI-l Pod (?) MI 5.9; pv. phaseolicola Pp-MI-B Velvet leaf weed MI 2.9_ pv. phaseolicola Pp-CH-Zl ? Malawi 5.99 pv. nhaseolicola Pp-HB-38 ? Phalombe.Malawi 5.9; pv. phaseolicola Pp-P23 Leaves (?) Emali. Kenya 5.9; pv. phaseolicola Pp—P8 Leaves (?) Arusha.Tanzania g.§_ pv. syringae Ps-MI-l Leaves (?) MI Corynebacterium michiganense Cm-MI-l Tomato leaves MI 9; fascians Cf-l Laboratory stocks Nebraska Agggbacterium tumefaciens At-l Tomato stem galls MI Erwinia amylovora Ea-l Laboratory stocks MI Bacillus megaterium Bm-l Laboratory stocks MI aMDA - Michigan Department of Agriculture (Seed lot number) bCultures were kindly supplied by Dr. A.K. Vidaver, Department of Plant Pathology. University of Nebraska. Lincoln. NE. 14 of 0.01M phosphate buffer (pH 7.2) to the lyophilized culture, lOOpfuls of the resulting suspension were streaked on YCA and incubated as previously described. Genetic stability and purity were routinely checked by careful morphological examination, repeated plating from dilute suspensions of cells and by pathogenicity tests. Pathogenicity tests. Bacterial suspensions were prepared from 24-hour-old cultures growing on YCA. Suspensions prepared in 0.01 M phosphate buffer (pH 7.2) were adjusted to optical density 0.2 at 620 nm, and contained 1.5 - 2.6 x 106 colony forming units (CFU) per milliliter. Fourteen-day-old greenhouse grown Seafarer plants were inoculated on the first trifoliolate leaf by water-soaking an area of 10 mm diameter using a 3 cc hypodermic syringe. Each isolate was replicated four times using one plant per replicate, and pathogenicity was evaluated 14 days after inoculation with plants kept at temperature ranges of 23-30 C in the greenhouse. Test media. Media were evaluated for selective inhibition Of microorganisms other than XCP that frequently cause problems in isolation, and for efficiency of recovery of the target organism from different sources. Initially, media used included the following: MXP -— This medium was prepared following the procedures described by Claflin 99 9I. (14) and contained (g/l): Monopotassium phosphate (KHZPO4) (0.6); dipotassium 15 phosphate (KZHPO4) (0.8); soluble potato starch (8.0); yeast extract (0.7); potassium bromide (10.0); glucose (1.0) and bacto-agar (15.0). Following autoclaving and after the medium had cooled to about 40—45 C in a water bath, the following antibiotics were added. One ml each of Chlorothalonil (Bravo) (1.2 ml in 38.8 ml water); Kasugomycin hemisulfate (200 mg in 100 ml water); Cephalexin monohydrate (200 mg in 100 ml of water) and gentamycin sulfate (20 mg in 100 ml water) at the rates 15, 20, 20 and 2 micrograms per milliliter, respectively. Glass distilled water was used to prepare stock solutions. Dyes included (microliters/L); 6.0 methyl green (1% aqueous solution) and 3.0 methyl violet 23 (1% solution in 20% ethanol). Dyes were incorporated into the medium before sterilization for 20 minutes at 250F (21). SSM: One gram yeast extract and 15.0 g bacto-agar were added to 970 ml of 0.01 M phosphate buffer (pH 7.2) and sterilized by autoclaving for 20 minutes at 250 F. After cooling to 45 C in a water bath, antibiotics were added aseptically to contain per liter; cycloheximide 25.0 mg; nitrofurantoin 2.0 mg; nalidixic acid 1.0 mg, and gentamycin sulfate 0.05 mg. The protocols for preparation of stock solutions were as outlined by Trujillo and Saettler (76). DSX—-A soluble starch medium developed by Dhanvantari (18) for isolation of Xanthomonas campestris pvs. campestris and phaseoli; contained soluble potato starch 10.0 g; yeast 16 extract 5.0 g; NH4H2PO4 0.5 g; MgSO4 7H20 0.2 9; NaCl 5.0 g; agar 15 g and Tween 80, 10.0 ml; in one liter of glass distilled water. The pH was adjusted to 6.8 and after autoclaving the antibiotic pimaricin (50 mg/l) was added. The medium was stored for 9 days at 5 C prior to use. The DSX medium was eliminated after initial evaluations since recovery of XCP was very low and the plates required a long time (9 days or more) of storage before use. All reagents and antibiotics were purchased from Sigma Chemical Corporation St. Louis, MO except cycloheximide which was purchased from Cal. Biochem. Behring Corp., La Jolla, CA. Stock solutions of antibiotics were prepared every two weeks, and after filter sterilization, stocks were stored in the refrigerator at 5-7C. Before use stock solutions were allowed to stand at room temperature or in the hood chamber for about one hour to reduce the temperature gradient between stock solutions and the media. In all cases, media were subject to constant stirring with a sterile magnetic stir-bar during addition of antibiotics; and after addition, stirring continued for about 3 minutes before dispensing into plates. This was deemed necessary to ensure uniform distribution of antibiotics in the media. The medium YCA (Yeast extract calcium carbonate agar) was used as a standard for comparison purposes. It contained per liter of distilled water; yeast extract 10.0 9; calcium carbonate 2.5g; bacto-agar 15.09. The medium was 17 steamed for 15 minutes and thereafter, 250 ml dispensed in 500 ml capped bottles, followed by sterilization by autoclaving as described elsewhere. These procedures were followed to avoid the heavy white precipitate that usually accumulates at the bottom of the culture plates causing problems during colony counts when using indirect light and when plates are counted without opening. Development of M-SSM medium. In the process of developing an improved semi-selective medium (M-SSM) agar was added in order to make SSM a solid rather than a liquid medium. Other changes included addition of dyes, soluble potato starch as a carbon source, and changing the concentration of gentamycin sulfate. To reduce number of background organisms, the concentration of gentamycin sulfate was increased from 0.05 microgram per milliliter to 0.5 microgram per milliliter. The amount of soluble potato starch varied from 8-10 9, but the final amount used was 8.0 g. Methyl green and methyl violet 23 were added at the rates of 6.0 microliters per liter and 3.0 microliters per liter, respectively. Attempts to incorporate glucose as another carbon source were made. However, it was finally omitted because XCP colonies became very slimy and difficult to handle during counting when glucose was incorporated into the media. The effect of these changes upon plating efficiency, 18 recovery, growth rate, genetic stability of XCP, color, differentiation of target organism from others; and the relationship between age of the poured medium and toxicity reported in other pathogen-media systems (19, 45, 50) relative to other media were also tested. Genetic stability of XCP recovered on these media was checked by observing colony characteristics and routine pathogenicity tests using 14-21 day-old Charlevoix bean seedlings. Pure culture studies. Early log phase cultures of various bacterial strains (Table l) were prepared by growing fresh cultures on YCA at 27 i 1 C for 24—28 hours. Bacterial suspensions were prepared by washing culture plates with 0.01 M phosphate buffer (pH 7.2) and adjusting suspensions to an optical density of 0.1 at 620 nm as measured with a Spectronic 20 colorimeter (Bausch and Lomb Company, NY). The suspensions were ten-fold serially diluted in the same buffer and decimal aliquots from different dilutions plated on the test media in triplicate. Aliquots were spread evenly on each plate with an L-shaped glass rod. For each medium, a different spreader was used; and Spreaders were sterilized between use during the experiment by immersion in ethanol (95%) and flaming. Plates were incubated as previously described and colonies counted after 2-5 days. The morphology of XCP growth on each medium was carefully noted for future identification from mixed populations. The time required for colonies to 19 develop, colony shape, color and size on the test media were also noted. Growth rate was studied by measuring colony diameter five days after plating. Plates containing 20-40 colonies were used, and for each isolate a total of 30 colonies were measured, ten colonies per replicate. Plating efficiencies were expressed as the percentage of colony forming units on semi-selective media relative to YCA. Recovery from naturally infected bean leaves. Naturally infected bean leaves bearing characteristic blight lesions were collected from various farms in Michigan (Saginaw, Bay City; Crop and Soil Science, and Botany Research farms at MSU, East Lansing) during 1985 and 1986 growing seasons. Samples were stored in the refrigerator until used. Two 10 mm-diameter discs were excised from individual lesions using a flame-sterilized cork borer. The discs were comminuted in a steam-sterilized mortor and pestle in 1.0 ml 0.01 M phosphate buffer, pH 7.2. The macerates were allowed to stand for two hours at room temperature followed by decimal dilutions in the same buffer. Seven dilutions were made and aliquots of 0.1 m1 of the last three dilutions were plated on each test medium in triplicates and incubated at 27 i 1 C for 5-8 days before counts were made. Experiments were repeated four times. Colonies that resembled XCP and that cleared starch were transferred to YCA and confirmed to be XCP by pathogenicity 20 tests using procedures described elsewhere. Recovery from artificially inoculated bean tissue. Bean plants (cultivar Charlevoix) were grown in the greenhouse in 10 cm—diameter clay pots containing a 3:1 mixture of soil/vermiculite, respectively. Plants were watered twice a day using tap water and air temperatures fluctuated between 24 and 30C. The eight isolates of XCP used in this study were recovered from lyophilized stocks. Before the experiment, portions of the lyophilized bacteria stored at room temperature in the dark were suspended in sterile phosphate buffer (0.01 M, pH 7.2) and the resulting suspensions streaked on YCA and incubated at 27 i l C. The resulting single colonies were transferred again to YCA and incubated at the same temperature for 48 hours. Bacterial suspensions were prepared in the same buffer and adjusted to 1.1 - 2.3 x 107 CFU/ml as estimated by dilution plating procedures. The first fully-expanded trifoliolate leaf on l4-day—old plants was spray-inoculated with bacterial suspensions to run-off using an atomizer attached to an electrically driven pump; air pressure was 12 p.s.i. After inoculation, plants were maintained in the greenhouse. Typical common blight symptoms were observed 8-10 days after inoculation. Pathogen recovery was performed by excising two 10 mm- diameter discs and macerating the discs in 1.0 ml of phosphate buffer as previously described. After sitting for 21 one hour, suspensions were serially diluted and decimal portions plated on test media in triplicates. Plates were incubated as before and colony counts made after 5 days; YCA was used as.a control medium. Recovery of XCP from blossoms of artificially inoculated bean plants. Bean plants, line I-84100, were grown to the two true leaf stage in the greenhouse as described elsewhere, and were spray—inoculated with XCP (MI— 3 isolate) using a bacterial suspension containing 3.1 x 107 CFU/ml in 0.01 M phosphate buffer, pH 7.2. At flowering (16 days after inoculation) five open blossoms were removed and comminuted in 10 ml of the same buffer using mortar and pestle. The macerates were left to stand for one hour followed by ten-fold serial dilutions. One-tenth ml of each dilution was spread over culture plates to check for presence of XCP. Plates were incubated 5 days at 27 1 1 C, at which time counts were made. Each dilution was replicated three times. Percent recovery was expressed as the mean number of CFU on semi-selective media relative to YCA times one hundred. Data were log transformed before analysis. Recovery of XCP from dry bean tissue powder. Naturally infected bean leaves collected from farmers' fields in Michigan were dried and ground to a fine powder using a sterile mortar and pestle. The powder was stored in sterile 22 glass test tubes at room temperature in the dark for 9 months before assay. After nine months, one-tenths-gram of the powder was added to 10.0 ml of sterile 0.01 M phosphate buffer, and thoroughly vortexed at medium speed. The suspension was extracted for one hour at room temperature with constant shaking on a horizontal shaker adjusted to 75 x 1.5 inch strokes per minute. Following 30 minutes settling in a transfer chamber, the suspension was serially diluted in the same buffer and decimal portions plated on the test media in triplicate. The plates were incubated at 27 i 1 C, and final colony counts were done after 5 days. Growth of other bacteria was also recorded. Four experiments were conducted using the same procedures. Recovery of XCP from bean dust collected from seed. One gram dry weight of been dust was obtained from bean seed grown in Puerto Rico by sieving. The dust was then mixed with 10 m1 of 0.01 M phosphate buffer, pH 7.2, in a sterile test tube and vortexed for one minute at high speed. After 5 minutes of debris settling, the suspension was ten-fold serially diluted. Aliquots of 0.1 ml were spread on the media in triplicates and plates were incubated as described elsewhere. Colony counts were done five days after plating and the extent of contamination observed. Two experiments were conducted Effect of media age on recovery of XCP. Media were prepared following protocols described previously and stored 23 in closed plastic bags at room temperature in the laboratory for 4, 15, 30 and 45 days. One-tenth-gram of naturally infected bean powder used in previous experiments was mixed with 10 ml of 0.01 M phosphate buffer, pH 7.2, and vortexed for one minute at high speed. The suspension was extracted for one hour at room temperature with constant shaking as described elsewhere. Following decimal dilutions in the same buffer, 0.1 ml aliquots were spread on the test media in triplicate and incubated for five days at 27 i 1 C. Percent recovery was expressed relative to YCA, the standard medium. Data were analyzed statistically with MSTAT program version 4.0 (Department of Crop and Soil Science, Michigan State University, East Lansing, MI) as a split-plot design with media as whole-plot factor and age as a subplot factor. Significant differences between means were estimated using Duncan's multiple range test. Recovery from artificially infested soil. A composite soil sample was prepared by mixing equal volumes of soil from four agricultural fields, all of which had no history of been production. A one-gram sample was mixed with ten milliliters of 0.01 M sterile phosphate buffer, pH 7.2, and the mixture was shaken for one hour on a horizontal shaker (7.5 x 1.5-inch strokes/minute) at room temperature. Bacterial suspensions of various XCP isolates were prepared from 48-hour old culture grown on YCA, using the same 24 buffer, and adjusted to an optical density of 0.15 at 620 nm wavelength. One ml of each bacterial suspension was added to the soil suspensions (separately) and the mixture was thoroughly shaken for one minute using a deluxe mixer set at high speed. The suspensions were allowed to settle for five minutes followed by decimal serial dilutions in the same buffer. Aliquots of 0.1 ml were spread on test media. To estimate the amount of inoculum added to the soil suspensions, one ml of the original bacterial suspension was added to ten milliliters of sterile phosphate buffer (same buffer as above), mixed as previously described, diluted and plated in the same manner on YCA. Colony counts were done 2-5 days after incubation at 27 i l C. The extent of contamination for each medium was also recorded. Representative colonies of XCP recovered were tested for pathogenicity using Charlevoix as a susceptible cultivar, following the same procedures described elsewhere. 25 RESULTS Pathogenicity tests of XCP isolates. Isolates of XCP used for evaluation of media were tested for pathogenicity using greenhouse-grown Seafarer bean plants. All isolates were pathogenic on bean plants (Table 2). Typical blight symptoms were observed 7-9 days after inoculation. Media evaluated. Semi-selective media evaluated in this investigation were MXP, SSM, DSX and M-SSM. The latter was developed as an improvement of SSM. The standard YCA medium was used as a control. Media were evaluated relative to ease of preparation, rapid growth rate of XCP, high plating efficiency, selectivity against non-target bacteria, recovery from various infected and infested materials, and effect of age on pathogen recovery. Of the semiselective media studied, M-SSM and SSM were found to be easier to prepare than MXP and DSX, in terms of time involved. The DSX medium was eliminated after initial evaluations, since recovery of XCP was very low and the plates required storage (minimum of 9 days) before use. Furthermore, recovery of XCP on DSX from naturally infected bean tissue was found to be significantly lower (P = 0.05) than MXP, SSM and YCA over four experiments conducted (Table 3). 26 Table 2. Pathogenicity tests of Xanthomonas campestris pv. phaseoli (Xp) and fuscans variant (prg isolates used for evaluation of selective media 'Isolate Host Reactionb MI—l (pr) MI-2 (Xp) MI—3 (Xp) MI-4 (Xp) MI-5 (Xp) MI-6 (pr) MI-7 (Xp) MI-8 (Xp) MI-9 (Xp) MI-10 (Xp) MI-ll MI-12 MI-13 MI-14 MI-15 MI-16 MI-l7 (Xp) (Xp) (Xp) (Xp) (pr) (pr) (Xp) Sc-4A (Xp) LB 2 (Xp) ++-+-++-++-++-+-++-++-++-+-++ aBacterial suspensions were prepared from 24-hour—old cultures growing on YCA. Suspensions were prepared in 0.01 M phosphate buffer (pH 7.2) to contain 1.5 - 2.6 x 10 CFU/ml. Fourteen day old green house grown Seafarer plants were inoculated by water-soaking a leaf area of 10 mm diameter. Each isolate was replicated four times and pathogenicity was evaluated 14 days after inoculation. One plant was used per replicate. b+ = Pathogenic, development of necrotic lesions surrounded by chlorotic areas. 27 Table 3. Recovery of Xanthomnas campestris pv. pliaflli; frun naturally infected bean leaves on MXP, SSM, DSX, and Mediumy Colony forming units/cm2 infected tissue Exp 1 Exp 2 Exp 3 Exp 4 meanz MXP 2.4x108 3.8x108 1.5x108 3.42:108 2.8x108 b SSM 1.2x109 4.911108 4.2x103 2.6x1o8 5.9x108 a sz 2.7x105 2.6x105 2.1x105 2.5x106 8.1x105 c YCA 1.3x109 4.6x108 2.2x108 3.7x108 5.9x108 a xNaturally infected bean leaves were collected from Saginaw, Michigan; and from Crop and Soil Science, and Botany farms at Michigan State University, East Lansing. Two-10 nm diameter discs containing equal amounts of healthy and infected tissue were removed fmn infected leaves and ccmninuted in 1 ml of 0.01 M phosphate buffer (pH 7.2). Suspensions were allowed to stand for 2 I'DUIS followed by ten-fold serial dilutions, 0.1 ml samples were plated on surface of plates of test media in triplicate. were made after 5-8 days. yRecipes for media are described in the text. Plates were incubated at 27 3: 1C, counts zMeans followed by the same letter are not significantly different according to mincan's nultiple range test at 5% level. 28 Time required for colonies to develop. On SSM, small XCP colonies appeared 2-3 days after plating; at about 4-5 days, colonies reached a maximum diameter of 1.7 - 3.6 mm depending on strain. The colonies were circular in form, convex with entire margins and smooth surfaces. Colonies appeared light yellow, and were not slimy (Figure l). Colonies of XCP were visible after 3 days on MXP, when zones of starch hydrolysis were observed around colonies. Starch clearing differentiated pathogen colonies from other microbes. XCP colonies were yellow in color, slimy, circular in form, convex with entire margins and smooth surfaces (Figure 1). After 5 days at 27 1 1C colonies ranged from 1.3 - 6.30 mm in diameter, depending on the strain used. On DSX, most XCP colonies were visible in 4-5 days and reached a diameter of about 3 mm in six days after plating at 27 1 1C. Colonies were yellow, smooth initially but changing to irregular shape with time. Zones of starch hydrolysis were produced, but the zones were weaker compared to those on other media. Colonies of the pathogen were visible on YCA in 3-5 days depending on strain and source of inoculum; colonies reached a diameter of 2-4 mm at 5 days. The colonies were yellow, circular, convex with entire margins. 29 Figure 1. Comparative growth and colony characteristics of Xanthomonas campestris pv. phaseoli on plates of SSM, MXP, M—SSM, and YCA (control), after five days incubation at 27 1 1C. 30 Development of improved semi-selective medium (M—SSM). Disadvantages of liquid semi-selective media are obvious. Numerous non-target organisms can grow without easy differentiation from the desired pathogen. For this reason, the liquid medium SSM was modified by adding agar (15.09) and soluble potato starch (8-10 g). The concentration of gentamycin sulfate was increased from 0.05 to 0.5 micrograms per milliliter to reduce background contamination. Methyl green and methyl violet 28 were added at 6.0 and 3.0 microliters, respectively, to facilitate visibility of zones of starch hydrolysis. Starch hydrolysis allowed distinguishing the pathogen from other background bacteria. With SSM, it was difficult to differentiate XCP from other saprophytic organisms. Addition of glucose (0.5 - 1.0 9) resulted in very slimy XCP colonies; for these reasons, glucose was eventually omitted. Increasing levels of soluble potato starch up to 109 did not improve recovery, hence the lower amount (8.09) was used. The final improved M-SSM medium contained: yeast extract 1.09; soluble potato starch 8.09; bacto-agar 15.09; methyl green (1% aqueous solution) 6.0 microliters; methyl violet 2B (1% solution in 20% ethanol) 3.0 microliters and 970 ml of 0.01M phosphate buffer, pH 7.2. After autoclaving and cooling to about 45C in a water bath the antibiotics were added aseptically to contain per liter; cycloheximide 25.0 mg; nitrofurantoin 2.0 mg; nalidixic acid 1.0 mg and 31 gentamycin sulfate 0.5 mg. On M-SSM, colonies of XCP appeared after 36—48 hours reaching a diameter of about 0.5 - 1 mm in diameter at 27 i 1C. At this time, zones of starch hydrolysis were easily seen around each colony under indirect light. Similar to MXP, zones of starch hydrolysis increased in size with increase in colony diameter, coalesced and disappeared when numerous colonies were present on the plate. In five days, colonies of the pathogen reached a diameter of 3.1 - 7.5 mm depending on strain used. Colonies of XCP on M—SSM were circular with entire margins, convex, light yellow at first but became yellow and slimy with age (Figure 1). The brown pigment- producing variant, 59 phaseoli var. fuscans did not produce pigment in M-SSM. Similar observations were made on SSM and MXP. Plating efficiency. Table 4 shows results of comparative plating efficiencies of several XCP isolates in pure culture. All isolates tested grew on SSM and M-SSM. Results obtained with MXP revealed that, of 19 pathogen strains tested, two fuscans isolates (MI-1 and MI-15) and one XCP isolate did not grow on the MXP. Plating efficiencies on MXP, SSM, and M-SSM ranged from 0-96%, 63- 111% and 54-125% with averages of 53%, 88% and 85%, respectively. Except for rate of growth, colonies of XCP on MXP were similar morphologically to those on M—SSM. 32 Table 4. Plating efficiency of various isolates of Xanthomonas campestris W (Xp) and fuscans variant (pr) on MXP, 88b4, arfl M-S 0 Plating efficiencyy Isolate MXP SSM NPSSM MI - 1 (pr) o.o az 79 b 73 b MI - 2 (Xp) 59 c 89 a 74 b MI - 3 (Xp) 88 a 91 a 97 a MI - 4 (Xp) 89 a 86 a 93 a NE:- 5 (Xp) 96 a 87 a 92 a MI - 6 (pr) 15 b 63 a 54 a Na - 7 (Xp) 0.0 a 90 b 80 b NE - 8 (Xp) 50 b 93 a 95 a MI - 9 (Xp) 87 b 111 a 91 b MI - 10 (Xp) 78 a 88 a 76 a MI - 11 (Xp) 2 b 72 a 94 a ME - 12 (Xp) 48 b 75 a 93 a MI - 13 (Xp) 66 b 103 a 88 a MI - l4 (Xp) 3 a 63 b 54 b NE — 15 (pr) 0.0 a 106 b 125 b MI - 16 (pr) 77 a 88 a 84 3 MI — l7 (Xp) 91 b 100 a 101 a Sc-4A (Xp) 81 a 88 a 85 a LB 2 (Xp) 80 ab 93 a 75 b xBacterial suspensions obtained frcmllog phase YCA cultures were adjusted to optical density 0.1 at 620 nm; tenrfold serial dilutions were made in 0.01 M phosphate buffer (pi-I 7.2), and 0.1 ml portions plated on test media and.incubated at 27.: 1 C. colonies were counted after 5 days. Recipes fer media are described in the text. YPlating efficiency = mean number of CPU on semi-selective medium/mean number of CFU’on‘YCA x 100. Three replications were used for eadh medium zWithinrows, numbers followed by the same letter are not significantly different by Duncan's multiple range test at 5% level. 33 Growth rates of various isolates of XCP were estimated by colony diameter measurements (Table 5). Approximately 92% of the isolates studied grew significantly faster (P=0.05) on M-SSM than on the other media. Colonies of the fuscans variant seemed smaller in size on MXP, and in one case, on M-SSM. Growth rates of most isolates on SSM were not significantly different from those on YCA; and for most isolates, colony size remained smallest of all. A number of bacterial species from other genera were tested for ability to grow on semi-selective media; all isolates tested failed to grow on MXP (Table 6). Included were six isolates of Pseudomonas syringae pv. phaseolicola, which also did not grow on M-SSM. However, 99 syringae pv. syringae (Ps-MI-l) grew both on SSM and M-SSM, but colony forming units were only 21% and 5%, respectively, of that observed on standard YCA. Corynebacterium fascians (Cf—1) and Erwinia amylovora (Ea-1) also grew at very reduced rates on SSM and M-SSM. Compared to YCA, growth of Cf-l on SSM and M-SSM media was only 5% and 4%, respectively. Bacillus megaterium grew on SSM only. In all cases, growth of these organisms was slower on the semi-selective media. No zones of starch hydrolysis were observed; as contrasted to XCP, where colonies always produced strong zones of starch hydrolysis (Figure 2). This indicated that the bacterial strains were not able to utilize starch. Moreover, colonies of non-target bacteria were smaller compared to those on 34 Table 5. Comparative colony development of Xanthomonas campestris pv. phaseoli (Xp) and fuscans variant (pr) on four test media five days after platingx Mean colony diameter (mm)y Isolate MXP SSM M-SSM YCA MI-l (pr) o.o dz 1.9 c 6.9 a 2.8 b MI-2 (Xp) 6.3 b 2.9 c 7.1 a 3.0 c MI-3 (Xp) 5.5 a 3.4 c 5.4 a 4.0 b MI-4 (Xp) 4.7 b 2.0 d 6.1 a 3.8 c MI—7 (Xp) 0.0 d 1.9 c 6.8 a 2.3 b MI-8 (Xp) 6.0 a 3.6 c 5.0 b 4.0 c MI-9 (Xp) 2.5 b 1.9 c 6.4 a 2.5 b MI-10(Xp) 3.8 b 1.8 c 7.5 a 2.0 c MI-12(Xp) 1.3 d 1.9 c 6.3 a 2.6 b MI-13(Xp) 1.6 c 1.9 c 6.1 a 2.7 b MI-15(pr) 0.0 c 1.9 b 6.0 a 2.1 b MI-16(pr) 1.6 c 1.7 c 3.1 b 3.9 a MI-17(Xp) 2.8 b 1.7 d 5.2 a 2.5 c xThe same plates as in Table 4 were used. yMeans of three replications, ten colonies were measured per replicate. Plates containing 20-40 colonies were used. Recipes for media are described in the text. zWithin rows, numbers followed by the same letter are not significantly different according to Duncan's multiple range test at 5% level. Table 6. 35 Recovery of various bacterial isolates on MXP, SSM, M-SSM and YCAa Isolate identification colony forming units per plate b number MXP SSM M-SSM YCA Pp-MI-l 0.0 0.0 0.0 176.3 Pp-MI-3 0.0 0.0 0.0 39.0 Pp-CH-21 0.0 0.0 0.0 34.0 Pp-HB-38 0.0 0.7 0.0 24.3 Pp-P23 0.0 0.0 0.0 103.0 Pp-P8 0.0 0.0 0.0 88.3 Ps-MI-l 0.0 6.7 1.7 31.3 Cm-MI-l 0.0 0.0 0.0 37.0 Cf-l 0.0 3.3 2.7 65.3 At-l 0.0 0.0 0.0 198.3 Ea-l 0.0 69.3 2.0 248.7 Bm-l 0.0 23.3 0.0 55.5 aBacterial suspensions were prepared from log phase YCA cultures to optical density 0.1 at 620 nm and were ten-fold serially diluted in 0.01 M phosphate buffer (pH 7.2). Aliquots of 0.1 ml were plated on test media in triplicate _ Colonies were counted after 5 Recipes for media are described in the text. and incubated at 27 + 1 C. days. bMean colony counts of three replicates for each medium. 36 Figure 2. Starch hydrolysis zones around colonies of Xanthomonas campestris pv. phaseoli on plates of M—SSM after four days of incubation at 27 + 1C. 37 YCA, and required longer times (3-4 days) to develop. Non- target colonies were white and flat, and easily distinguished from the pathogen. Recovery from naturally infected bean leaves. Media were evaluated for efficiency of pathogen recovery from naturally infected bean tissues homogenized in 0.01M phosphate buffer, pH 7.2. Over the four experiments conducted, XCP was recovered in varying levels on these media. Percent recovery on MXP, SSM and M-SSM of XCP from naturally infected bean leaves is shown in Table 7. Few significant differences (P=0.05) in percent recovery were found between the three media within experiments. In general, the mean percent recoveries for the four experiments were not significantly different. Pathogenicity tests of presumed XCP colonies were positive. Except for YCA, saprophytic growth of non-target bacteria was not observed. Recovery from artificially inoculated bean tissue. Recovery of XCP from artificially inoculated bean leaves was estimated using greenhouse grown Charlevoix plants. Comparisons of recovery of various XCP isolates from artifically inoculated bean tissue on the test media are given in Table 8. The rate of colony appearance on all media was similar to that of pure cultures. Contaminants did not interfere with counts on semi-selective media. Even 38 Table 7. Recovery on three semi-selective media of Xanthomonas campestris pv. phaseoli from naturally infected bean leaves.2L Percent 59 E; pv phaseoli recovered ony Experiment MXP SSM M-SSM 1 71.6 cz 102.5 a 87.9 2 104.0 a 115.3 a 94.7 3 84.9 b 91.0 b 80.4 4 67.2 b 95.0 a 89.2 Mean 81.9 a 100.9 a 88.1 xNaturally infected bean leaves were obtained from farmers' fields in Michigan. Tissue was comminuted in sterile 0.01 M phosphate buffer (pH 7.2). Suspensions were allowed to stand for 1 hour before ten-fold serial dilutions were prepared, from which 0.1 ml portions were plated on each medium in triplicate; plates were incubated at 27 i 1 C for 5 days. yPercent recovery = recovery on semi-selective medium/recovery on YCA x 100. Recipes for media are described in the text. zNumbers followed by the same letter within rows are not significantly different by Duncan's multiple range test at 5% level. 39 Table 8. Recovery on four test media of various isolates of Xanthomonas campestris pv. phjgeoli from artificially inoculated bean leaves“ Colony forming units per platey Isolate MXP SSM M—SSM YCA 141-2 224.3 abz 197.2 b 267.0 a 199.0 b MI—4 94.3 a 102.7 a 100.7 a 91.3 a MI-8 181.3 a 153.0 b 187.7 a 175.7 a MI-9 215.0 b 260.3 a 263.3 a 274.0 a MI-10 80.6 ab 82.0 ab 86.7 a 75.7 b MI-ll 248.0 b 290.3 a 270.3 ab 297.0 a MI-13 152.3 a 166.0 a 172.7 a 175.0 a MI—17 91.7 a 112.7 a 94.7 a 100.0 a xBean plants (cultivar Charlevoix) were spray-inoculated when the first trifoliolate leaf was fully expanded (14 days after planting) with bacterial suspensions containing 1.1 - 2.3 x 10 CPU/ml. Tissue samples were taken 14 days after inoculation when disease symptans were fully developed. Two-10 mn diameter discs were renoved frtm infected tissue and comminuted in 1.0 ml of 0.01 M phosphate buffer (pH 7.2). After standing for one hour, suspensions were serially diluted and 0.1 m1 portions plated on test media in triplicates. Plates were incubated at 27 1 1C and colony counts were made after 5 days . yValues are means of three replications. Recipes for media are described in the text. ZWithin rows, means followed by the same letter are not significantly different by duncan's multiple range test at 5% level. 40 on YCA, contamination was minimal, suggesting that saprophytic populations were diluted out. Generally, colony counts of XCP on different media did not very much. Few significant differences in colony counts (P = 0.05) were observed (Table 8). Recovery from blossoms of artifically inoculated plants. Colony counts were significantly lower (P=0.05) on MXP than the other media (Table 9). Recovery on M-SSM was 9% less than that on SSM. Numbers of colony forming units on SSM and M-SSM media were comparable to that on YCA. However, of the semi-selective media, SSM gave the highest recovery. Recovery of XCP from dry tissue samples. Xanthomonas campestris pv. phaseoli was readily recovered from extracts of naturally infected bean leaf powder stored at room temperature for nine months (Table 10). Over the four experiments conducted, colony counts on MXP were significantly lower (P=0.05) than on other semi-selective media. however, MXP was 100% effective in reducing saprophytic populations of both bacteria and fungi, followed by M-SSM. Of the semi-selective media studied, SSM supported the highest levels of non-target bacteria. On YCA, presence of fast spreading saprophytic bacteria, in some cases, prevented recovery of XCP. Colonies of XCP consistently appeared on M-SSM within 48 hours. Development 41 Table 9. Recovery of Xanthomonas campestris pv.‘ m from flowers of artifiwa inoculated bean plants grovm in the green house.v MediumW CPU/five flowersx % recoveer MXP 8.3 1 0.6 x 103 bz 83.3 SSM 1.1 i 3.6 x 104 a 104.0 M-SSM 1.0 3: 4.2 x 104 a 95.6 YCA 1.1 i 2.3 x 104 a -- VPlants of line I-84100 were grown in the green house and were spray- inoculated with MI-3 isolate (3.1x10 CPU/ml) when possessing two trifoliolate leaves (18 days old). wRecipes for media are described in the text. xAt flowering (16 days after inoculation), five open flowers were calmintrted in 10 ml of sterile phosphate buffer (0.01 M, pH 7.2) usingamortarandpestle. Thesuspensicnwas lefttostandforone hour followed by ten-fold serial diluticns in the same buffer. Portionsof0.1mlwereplatedontestmediaandincubatedat27il C for 5 days. Values are means of three replications. YPercent recovery = mean number of colony forming units (CFU) on semi- selective medium/mean number of CFU on YCA x 100. zMeans followed by the same letter are not significantly different according to Duncan's multiple range test at 5% level. 42 Table 10. Recovery on four test media of Xanthomonas campestris pv. phaseoli from infected dry bean tissue stored for nine months at room temperature.w Colony forming units per platey Mediumx Exp 1 Exp 2 Exp 3 Exp 4 MXP 88.3 c 51.3 b 56.7 c 121.0 b SSM 255.3 ab 110.0 a 119.0 a 197.0 a M-SSM 273.3 a 92.0 a 111.0 ab 186.0 a YCA 231.7 b 0.0 cz 84.0 bc 155.7 ab wInfected bean leaves were dried at room temperature, ground into a powder and stored in sterile glass test tubes at room temperature for nine months. One-tenth-gram of the powder was added to 10 ml of sterile 0.01 M phosphate buffer, pH 7.2 and thoroughly vortexed at medium speed. The suspension was extracted for one hour on a horizontal shaker adjusted to 75xl.5 inch strokes/min. After settling for 30 minutes, the suspension was ten-fold serially diluted in the same buffer and decimal portions plated on the test media. Plates were incubated at 27 i 1 C for 5 days. xRecipes for media are described in the text. yMeans of three replicates for each test medium. Within a column, means followed by the same letter are not significantly different at 5% level by Duncan's multiple range test. zNo 59,99 pv. phaseoli colonies were recovered due to high numbers of contaminants. 43 of the pathogen on MXP and SSM was occasionally not uniform. Recovery of XCP from bean dust collected from seed. The mean number of colony forming units of XCP recovered on the test media and the level of contamination for the two experiments conducted are given in Table 11. Percent recovery was highest on SSM followed by M-SSM. Reduction of other microbial growth was highest on MXP, and least on SSM. In one experiment, recovery of XCP on YCA was hindered by high levels of saprophytic populations (Figure 3). Colonies of the pathogen were easily identified on MXP and M-SSM based on presence of starch hydrolysis zones, growth rate, colony morphology, and were confirmed by pathogenicity tests of representative colonies. Presumed XCP colonies on all media proved pathogenic when tested on bean plants. It is not known whether the pathogen recovered from dust resulted from contamination of the seed during handling of seed or during processing. Probably the seeds from which dust was collected were obtained from bean plants grown in infested fields. Effect of media age on recovery. Studies were conducted using media stored in plastic bags at room temperature for 4, 15, 30 and 45 days. This prevented excessive moisture which occurs when plates are stored in a refrigerator, causing difficulties in obtaining single colonies for counts. Assay was done using extracts from Table 11. Recovery of Xanthomonas campestris pv. 44 dust collected from seed grown in Puerto Rico eoli from bean Mean # of CPU/gram of bean dustc Medianb Exp# % Recoveryd 5._c_. weoli Contaminants MXP 1 6.0 _+_ 3.06x102 0.0 - 2 3.1 : 1.00x103 2.3 : 0.88x102 333.3 SSM 1 1.3 : 0.58x103 5.3 _+_ 0.67x102 - 2 3.6 i 2.40x103 1.6 : 3.0521103 390.3 M—SSM 1 1.5 : 0.331103 0.67: 0.33x102 - 2 3.4 : 2.191803 5.3 : 0.33x102 368.8 YCA 1 0.0 TN'I‘Ce - 2 9.3 : 2.03x102 4.6 : 7.13x103 - aOnegramdryweightofbeandustwasmixedwith10mlofphosphate buffer (0.01 M pH 7.2) and vortexed for one minute at high speed. Debris was allowed to settle for five minutes followed by ten-fold serial dilution. Portions of 0.1 ml were spread (11 the media in triplicate, plates were incubated at 27 i 1C for 5 days. bRecipes for media are described in the text. CMean colony counts of three replications followed by the standard error. dPercent recovery = mean number of CPU recovered on semi-selective medium/number of CFU recovered on YCA x 100. e'I'N'I‘C='I‘oonumeroustocount. 45 Figure 3. Recovery of 59n9n9n9n99 99np9999I9 pv. phaseoli on YCA (control), SSM, MXP and M-SSM from bean dust collected from seed grown in Puerto Rico. Photograph taken after 5 days incubation at 27 1 1C. 46 infected bean leaf powder. For all semi—selective media tested, percent recovery was unaffected by media age (Table 12). However, it was noted that time required for aliquots to be absorbed by media decreased progressively with age due to less water content in the media. Compared to SSM and M- SSM, percent recovery on plates of MXP was significantly lower (P=0.05) at 4 and 15 days. Recovery from artificially infested soil. Recovery of XCP and the level of contamination recorded after 5 days incubation at 27 i 1C are given in Table 13. Colony forming units added to the soil were estimated by diluting the same amount of inoculum (1.0 ml) in 10 ml of the same sterile buffer. For reduction of growth of other microbial organisms MXP was the medium of choice; it was superior to any of the media evaluated. Recovery of XCP on M-SSM was nearly the same as on MXP, however, levels of contamination were somewhat higher. Of the semi-selective media evaluated, SSM had the highest level of contamination; growth rate of contaminants was slower compared to YCA. Generally, fungal contamination was rare on the semi- selective media. Although soil was infested with large numbers of the pathogen, very few colonies of the pathogen were detected on YCA. In 40% of the cases, XCP could not be recovered on YCA from infested soil due to presence of fast growing fungi and bacterial contaminants (Table 13). 47 Table 12. Effect of nedia age on anmt recovery of Xanthoncnas campestris pv. pr_1as_e9I_i_ % recovery with media of age (days) Mediuny 4 15 3o 45 MXP 74.9 dz 66.2 a 77.2 d 82.3 c ssm 91.3 be 96.3 abc 82.3 d 93.6 abc M-SSM 89.9 bc 93.9 bc 85.7 d 88.5 bc WStorage was done in closed plastic bags at room temperature. Naturally infected bean leaf powder was used for assay. One-tenth gram of bean leaf powder was mixed with 10 ml of 0.01 M phosphate buffer (pH 7.2) and vortexed for one minute at high speed, and extracted for one hour. Following decimal dilutions in the sane buffer, 0.1 ml aliquots were spread on the test media in triplicate andincubatedforSdaysat27_+_1C. xPercent recovery = average number of CPU on each said-selective medium/average number of CPU on YCA x 100. YRecipes for nedia are described in the text . zP‘igures followed by the same letter are not significantly different according to Duncan's multiple range test at 5% level. 48 mfl.mmmmdmmdh0mmtflsw.wmmmm artjridanyixrestedsoiia Mam (EU/m1 Isolate ”(P $1 [VI-$4! YCA 141-2 157394.24 4.79.67 147.09.15 20.39.33 154.39.20 4.39.67 52.09.77 32.09.87 1790953005 141-3 136015.54 0.0 203.3922 11.39.19 207615.21 239201793981 20.79.38 2203953406 141-4 15639.28 2.39.67 143.7933 1.39.33 160.09.15 1.0_+_0.58 75.09.86 26.09.08 171.39.45m106 MIL-8 136.39.45 0.0 13209.04 3.79.33 13379.84 239201197963 20.39.67 126.79.76x106 141-9 241.092.12 0.0 244.3942 7.79.19 249.3929 0.39.33 26.09.08 mcc 257.09.91me 141—10 274394.17 0.0 285.0955 14.39.84 29505813 5.39.19 29.3939 mrc 3050921905 1411-12 2137.79.38 0.0 244.0941 9.09.73 249.3929 0.3929 34090.6 11m: 2527940805 141-13 142.7_+_4.67 0.0 153.7926 15.09.06 150.3999 6.3_+_0.33 7539121587928 154.39.3de5 141-16 48.09.08 0.0 60.39.88 8.39.08 56.09.65 0.79.33 11.19.77 'INIC 68.09.13906 141-17 254.0951 0310332377910 14.09.61 254716.77 5.79.45 20039.74 30.89.08 2827984906 aAmtpxitesoilsalplevmquaaredbyndydngsoilfmnfarmaumlfiekh.allofdddilndmhistayofbem m. Aaegransarplemsndxedwithtenndllilitersofflrqietelnffa (0.01M. 1147.2). ‘mendxturem mamwmamtalstflaflfixljmmmdntte)afll.0mlofbacta'ialaaperaian(m0.15 at6m)vmadied. Aftedfldmfathmte.tm—fddsaialdilmiaamnadeaflo.lmmatedmmtst nediunintriplimte. PlateswereirnbatedatflilcwmlmymntsmdneafterSdays. hIl'leslsmepma‘grimetoallrurioersineachrov. Recipafcrnecfiaarebcribedintlietm. qMC-Tbonmustomnt. 49 Representative colonies of XCP recovered were tested for pathogenicity and found pathogenic. DISCUSSION The effectiveness of any semi-selective medium for isolating a target organism partly depends on the nature of the materials from which isolation is made. Thus efficiency of recovery varies depending on the level, condition and the state of the pathogen, as well as the diversity and the level of other microbial organisms present. Obviously, it is very difficult to develop a semi-selective medium which will always be efficient under all situations. However, efforts should be directed towards minimizing levels of non- target organisms and maximizing levels of the required pathogen. In addition, the medium should allow quick and easy differentiation of the pathogen from other saprophytic bacteria. In the current investigation diverse sources of material were used for evaluation of test media so as to represent different ecological niches for a variety of microbes. XCP was easily and efficiently isolated from various infected bean material and from infested soil on M-SSM. In terms of growth rate of 19 XCP isolates tested, M-SSM was superior to both MXP and SSM. M-SSM also allowed growth of all XCP strains tested (Table 4). Identification of XCP 50 51 colonies on M-SSM was relatively simple based upon growth rate, colony morphology and color, and presence of starch hydrolysis zones (Figure 2). Although M-SSM allowed growth of some other non-target bacteria, these were easily differentiated from XCP by lack of starch hydrolysis zones, absence of xanthomonadin pigment production, and colony morphology. Growth rate of these bacterial contaminants was also delayed. After storage of plates in plastic bags for up to 45 days at room temperature, M-SSM was able to recover XCP as well as on freshly prepared plates. Similar observations were made for MXP and SSM (Table 10). Based on pure culture studies, SSM medium proved superior in plating efficiency over any other semi-selective media tested. However, overall results indicated that M—SSM may be used successfully in detection and isolation of XCP from sources containing low populations of the pathogen. The largest reduction of saprophytic growth was observed on MXP. One of the reasons for such observations is that MXP contains a higher concentration of gentamycin sulfate (2.0 micrograms per m1) than M-SSM (0.5 micrograms per ml). In addition, MXP contains cephalexin which inhibits Erwinia 9pp., especially 9. herbicola; and kasugomycin which is effective against Pseudomonas 9pp. and other contaminants (14, 44). The highest recovery of XCP observed on MXP in this study (96%) coincided with that reported by Claflin et a1 52 (14). However, the major drawback with MXP was that, some strains of XCP did not grow on the medium (Table 4). The high concentration of gentamycin sulfate may account for lack of growth of some XCP strains on MXP. In their dose- response studies, Trujillo (75) and Trujillo and Saettler (76) indicated that XCP isolates used were able to tolerate gentamycin sulfate up to 0.5 micrograms per milliliter, although growth expressed in optical density at 620 nm, was somewhat reduced at this level. XCP isolates used by Claflin et a1 (14) tolerated up to 3 micrograms per milliliter; growth was expressed in colony forming units per milliliter. With gentamycin sulfate at a concentration of 0.5 micrograms per milliliter, M-SSM in most cases, did not differ significantly from SSM in percent recovery of the pathogen. These discrepancies observed between the current investigation and the results of Claflin et a1 (14) and Trujillo (75) may be attributed by differences in test organisms and differential sensitivity to gentamycin sulfate of specific XCP strains used. Indications of differential sensitivity to antibiotics between XCP isolates was observed with fuscans variants, which tended to be smaller in colony diameter, especially on MXP (Table 5). These findings further emphasize the fact that no medium appears best for all strains of the pathogen (46). Differences in the media 53 used in these previous studies may also account for variation in results obtained with these semi—selective media. Fungal contaminants on M-SSM were inhibited by using cycloheximide (25 mg/l). Nalidixic acid was added to inhibit Gram negative, cocci bacteria; it acts by inhibition of DNA synthesis (74). Nitrofurantoin has a wide spectrum antibacterial activity, but does not affect xanthomonads, and was included to suppress growth of both Gram-positive and Gram negative bacteria. Corynebacterium fascians, Erwinia amylovora and Pseudomonas syringae pv. syringae grew on M-SSM at a very reduced rates. In addition to the previously mentioned bacteria, 59 syringae pv. phaseolicola (Pp-HB-38) and Bacillus megatarium grew on SSM (Table 6). Bacillus 9pp. are reported to be common contaminants on bean seed (75) and in soil (17). Randhawa and Schaad (54) also documented that the most common antagonists found on crucifer seeds are spore-forming Bacillus 9pp. that can survive for long periods. These bacterial contaminants are said to have antagonist activity for xanthomonads (53, 54) and can easily prevent the isolation of the required pathogen. Growth of Erwinia 9pp. and Pseudomonas 9pp. on M-SSM can further be suppressed by amendments such as Cephalexin (14, 44) and kasugomycin or tabromycin (14), respectively. Besides facilitating clear vision of starch hydrolysis zones, methyl violet 2B and methyl green also 54 inhibited many Gram positive bacteria (11, 21). Findings presented in this study suggest that M-SSM is a promising tool for epidemiological studies involving routine laboratory detection and isolation of XCP from various infected and infested material. CHAPTER TWO POPULATION TRENDS OF XANTHOMONAS CAMPESTRIS PV. PHASEOLI IN REPRODUCTIVE TISSUES OF DIFFERENT BEAN GENOTYPES IN THE FIELD 55 INT RODUCT I ON A better understanding of the means of survival and mechanisms of transmission of phytopathogenic bacteria is an important element in efforts to improve the control of these pathogens (9). Debris from diseased plants has always been a possible source for seasonal carryover of pathogens. Investigations have revealed that phytopathogenic bacteria can also survive in non-host plants (9, 35, 62). Schuster (63) and Yoshii (49) indicated that XCP and its fuscans variant, can survive between growing seasons in the temperate zones within infested bean debris. These authors also reported that survival occurs in leaves placed on top but not 20 cm below the soil surface, and that survival was greater under dry than moist environmental conditions. Sutton and Wallen (72) could not isolate XCP from soil in which infected plants had been growing. Other workers (64, 82) believe that survival of XCP in the tropics may be greater than in temperate zones because of the opportunities to continually increase populations and epiphytic survival on perennials. Recently, Saettler et al. (56) presented evidence that XCP does not overwinter in Michigan. These contentions were based on inability to isolate pathogenic 56 57 XCP from 191 infected crop debris samples over 6 years of study; and lack of typical symptoms of the disease on the subsequent bean crop. They however, suggested that, since little is known about the factors governing infection of healthy plants from overwintered inoculum, inoculum load might have been too low to initiate infection. Although other sources of primary inoculum have been identified, their significance in the epidemiology of common and fuscous blights is not well documented. Such findings further indicate that infected seeds remain the most important primary source of XCP inoculum (9, 56, 80, 83, 84). Earlier studies indicate that common blight bacteria have the ability to enter the pod through the vascular system and infect the seed without causing lesions on the surface of the pods (83). In entering seed through the vascular system the pathogen frequently causes only a small yellow discoloration at the hilum. This symptom is not readily detected in colored seeds; such is not the case in white seeded types, where normally there is a slight yellow marking about the hilum. Seeds that possess only a small amount of infection without outward symptoms may cause much damage when planted the following season (72, 83). It is therefore apparent that selection of pods without symptoms is not an adequate means of control because vascular seed infection is very important. 58 Foliage and stems of tolerant bean cultivars are known to harbor relatively high populations of plant pathogenic bacteria without exhibiting discernible symptoms (9, 80, 81). Thus, there is a fear of using seeds from resistant bean genotypes grown in infested areas. This is based on the findings that even resistant cultivars grown under such conditions may produce both infested and infected seeds. Cafati and Saettler (9) and Schuster et a1. (65) have demonstrated that seeds of resistant bean genotypes can become infected with XCP to the same degree as seeds from susceptible genotypes. However, such findings were based upon studies in which the sutures of pods were artificially inoculated. This investigation addresses the question of seed infection in resistant and susceptible bean genotypes when grown and inoculated under field conditions during the seedling stage of development. An improved semi-selective medium (M-SSM) was used as a detection tool for XCP. 59 MATERIALS AND METHODS Location of the experiment The experiment was conducted under field conditions at Botany and Plant Pathology research farm, Michigan State University, East Lansing, Michigan, during the summer of 1986. Bean genotypes Seven bean genotypes were chosen on the basis of their reported reactions to common blight bacteria. Five susceptible cultivars (Pinto UI—114, Cranberry Taylor Hort., Charlevoix, Black-Magic and C-20) and two resistant bean genotypes (Valley and I-84100--a breeding line) were used. Experimental design: A completely randomized design (CRD) with four replications for each bean genotype was used. Seeds were hand-planted on June 27, 1986, in single six meter rows with 50 cm between rows, and a spacing of about 6-9 cm between seed within a row. Disease free seed of all genotypes was used. Weed control was accomplished by both hand and mechanical weeding. During dry spells soil moisture was supplemented by overhead sprinkler irrigation. 60 Bacterial isolates and inoculum preparation Michigan strain (MI-17) of XCP isolated from infected leaves of C-20 cultivar, was used. Cells from 24-hour YCA (yeast extract 10.0 9; calcium carbonate 2.5 g; bacto-agar 15.0 g and glass distilled water 1000 ml) cultures incubated at 27 i 1 C, were washed into sterile flasks with sterile phosphate buffer (0.01 M, pH 7.2), and diluted to an optical density of 0.2 at 620 nm on a spectrophotometer (Model Spectronic 20, Bausch and Lomb, Rochester, NY). Suspensions were then diluted further with the same buffer to a concentration of ca. 1.4 x 107 colony forming units (CFU) per milliliter as determined by dilution plating technique. The inoculum was used within 40 minutes after preparation. Inoculation of plants Bean plants were inoculated by gently spraying the foliage when the second trifoliolate leaf was fully open (18 days old). All the foliage was inoculated to run-off without water-soaking, using a Knapsack sprayer. Field evaluation of disease reaction. Disease reaction was evaluated progressively at ten, sixteen, and forty days after inoculation, respectively. For each disease rating date, plants were carefully examined for development of typical blight symptoms and the stage of plant development was noted. Pods were observed for 61 symptoms both during the growing period and at harvest time; pod reactions were noted at physiological maturity. In both cases, a CIAT scale of 0-9 was used; where: 0 = immune, no symptoms; 1 = 1% of leaf area covered with lesions; 3 = 5% of leaf area covered with lesions; 5 = 10%; 7 = 25%; 9 = 50% of leaf area or more covered with lesions. The extent of systemic development of the pathogen as revealed by the occurrence of symptoms in the upper uninoculated leaves was also noted for each bean genotype. Measurement of bacterial population. Studies of bacterial populations in reproductive tissues of various bean genotypes were performed at the flower bud, open flower, flat and bumpy pod stages of development. At flower bud formation, a sample of 25 flower buds was randomly removed from each replication using steam sterilized forceps (one forceps per replication) and placed in sterile 2.5 cm diameter glass test tubes. The same sampling procedure was used for open flowers. All samples were kept cold in the incubator until processing. Surface populations of XCP on flower buds and open flowers were estimated by shaking the samples for 30 minutes in 40 m1 of 0.01 M phosphate buffer containing 10 mM Magnesium sulfate and 0.01% Tween 20 (14) on a horizontal shaker set at 75 x 1.5-inch strokes per minute. After ten- fold serial dilution of the washates in the same buffer, 0.1 62 ml portions of each dilution were plated on M-SSM in triplicate and plates incubated at 27 i 1 C (unless otherwise stated, all incubations were done at this temperature). Colonies were counted over the range of five days of incubation. Counting was done on the back of the plates using a marker pen when viewed by transmitted light against a dark background. From plates of each genotype, ten representative XCP colonies were selected at random and tested for pathogenicity. For internal populations of XCP, samples of 25 flower buds and open flowers were surface sterilized in 2.62 percent NaOCl for three minutes and rinsed in three changes of sterile glass distilled water. Samples were then comminuted using sterile mortars and pestles in 30 ml of sterile 0.01 M phosphate buffer containing 10 mM Magnesium sulfate and 0.01% Tween 20. The macerates were allowed to stand for 15 minutes and decimal serial dilutions prepared, mixing well at each dilution; 0.1 ml aliquots were plated on M-SSM. Colonies were counted starting 2-5 days after incubation as previously described. Bacterial populations were expressed as CFU per gram fresh weight of tissue. Ten colonies from plates of each genotype were increased on YCA and tested for pathogenicity as described previously. In the field, blossoms were carefully examined for presence of symptoms since heavily infected plants experienced severe blossom drop. 63 Surface populations of XCP on flat and bumpy pods were evaluated by using six pods per replicate. Symptomless pods at both flat and bumpy stages were randomly sampled, placed in sterile plastic bags and kept cold until processed (24-36 hours), as described previously. Pods were shaken for 30 minutes in 60 ml of 0.01 M phosphate buffer containing 10 mM Magnesium sulfate and 0.01% Tween 20 Washates were decimally diluted in the same buffer and 0.1 ml aliquots of washates were plated on M-SSM and incubated for 2-5 days. Representative colonies of XCP from pods of each genotype were tested for pathogenicity. Colony counts were expressed as numbers per square centimeter of pod area. Pod area was measured using Lambda instrument, Model LI3000. Estimation of epiphytic bacterial contaminants associated with bean reproduction tissue. For studies of surface bacterial contaminants from flower buds, open flowers and pods, the same washates and procedure described previously for estimation of surface XCP were used. Aliquots of 0.1 ml were plated on YCA and incubated for 5 days before colony counts were made. Bacterial contaminants were differentiated from XCP based on growth rate, size of colonies, color on YCA, colony morphology and pathogenicity tests. On YCA, bacterial contaminants appeared within 24-48 hours while XCP required about 72 hours or more to appear. Single colonies of 64 bacterial contaminants were selected based on colony morphology and color on YCA and purified further by a series of transfers on the same medium. Purified isolates were maintained in 0.01 M phosphate buffer/glycerol (40%) mixture at 5-7 C for further studies. No attempts were made to further identify these bacterial epiphytes. Pathogenicity tests. Pathogenicity of XCP colonies cultured individually on YCA, was tested. Inoculum was prepared in sterile phosphate buffer (0.01 M, pH 7.2) to contain ca. 106 CFU/ml. Fourteen to twenty—one-day-old greenhouse-grown Charlevoix (kidney) bean seedlings were inoculated with bacterial suspensions by water-soaking an area of 10 mm diameter at four sites per leaflet. Each isolate was replicated four times using one plant per replicate. A known pathogenic XCP isolate and sterile phosphate buffer (same as above) were included as positive and negative controls, respectively. Plants were incubated in the greenhouse with temperatures ranging from 24 - 31 C, and observed on a daily basis. Pathogenicity was evaluated 14 days after inoculation. Appearance of typical common blight symptoms confirmed the identity of XCP. Assessment of seed infection. At physiological maturity, pods were hand-harvested from each replication and allowed to dry under greenhouse conditions. After drying, pods were shelled by hand and CC 011 fi 65 seed for each replicate kept separately. Seed samples were assayed for the presence of XCP by a modification of the procedures used by Cafati and Saettler (9). A random sample of one hundred seed was drawn from each replication using a slotted board with one hundred holes. Seeds were surface sterilized in 2.62 percent NaOCl for three minutes and then rinsed three times in sterile glass distilled water. Seeds were blotted dry on sterile paper towels and aseptically placed hilium down directly on the surface of improved semi- selective medium (M-SSM) containing 1% bacto-agar; five seeds per plate. Plates were incubated at 27 1 1C and observed daily for five days. Bacteria from infected seeds which produced a zone of starch hydrolysis on the medium were streaked on YCA and purified by a series of transfers. The identity of XCP was verified by pathogenicity tests as previously described. To further verify the absence of common blight bacteria, for each replication, 10 seeds showing no evidence of starch clearing in the medium were rechecked by a six-hour incubation (individually) in 4 ml of 0.01 M phosphate buffer containing 10 mM Magnesium sulfate contained in test tubes. The resulting soakates were spread on M-SSM and YCA and incubated as described elsewhere for five days. 66 Statistical analyses Data were analysed statistically with MSTAT program version 4.0 (Department of Crop and Soil Science, Michigan State University, E. Lansing, MI). Where appropriate, data were log normal and arcsine transformed before analysis (40). Significant differences between treatments were estimated using Duncan's multiple range test. 67 RESULTS Disease reactions. Disease reactions were evaluated progressively at ten, sixteen and forty days after inoculation. Foliage reactions of various bean genotypes used in this study are shown in Table 1. Under conditions of the experiment, the first foliage symptoms were observed 8 days after inoculation on all genotypes except I-84100 and Valley, on which symptoms appeared 15 days after inoculation. Initially symptoms appeared as water-soaked lesions on the inoculated leaves, which later coalesced to form large necrotic lesions surrounded by chlorotic areas. Lesions spread very rapidly within inoculated leaves on cultivars Cranberry Taylor Hort., Pinto UI-114, Charlevoix and C-20; lesions were accompanied by bacterial ooze in the form of shiny exudate on the abaxial surfaces of the leaves. A number of genotypes reacted similarly to the XCP isolate used. Resistant genotypes I-84100 and Valley were not significantly different from each other for the 3 disease ratings. Systemic movement of the pathogen as revealed by the occurrence of disease symptoms in the upper uninoculated leaves was very pronounced in Cranberry Taylor Hort., Pinto UI-114, and Charlevoix. At sixteen days after inoculation (blossom stage) the entire canopy on Cranberry 68 Table 1. Leaf reactions of seven bean genotypes to Xanthomonas canpestris pv. PM (MI-17) under field Disease reaction at various days after inoculationy Genotype 10 days 16 days 40 days I-84100 0.0 i 0.00 g2 1.4 _+_ 0.14 f 2.3 3: 0.29 ef Valley 0.0 i 0.00 g 1.5 i 0.24 f 2.5 i 0.33 f Pinto UI—114 4.5 i 0.33 d 7.8 i 0.29 b 6.5 i 0.33 c Cranberry 6.3 r 0.28 c 8.8 1 0.29 a 9.0 i 0.00 a Taylor Hort. Charlevoix 4.8 i 0.28 a 8.5 1 0.33 ab 6.3 1 0.55 c Black Magic 5.0 1 0.47 a 6.8 i 0.29 c 4.5 r 0.75 d c - 20 4.3 r 0.29 d 7.8 1 0.55 b 6.5 i 0.58 c inghteen—day—old bean pl ants were inoculated by gently spraying the foliage to run—off, without water-soaking, with a bacterial suspension cOntaining ca 1.4 x 10 colony forming units (CPU) per milliliter using a Knapsack sprayer. yMeans of four replications plus/minus standard error. Disease rating based on CIAT scale of 0-9; where 0 = inmune, no symptans; 1 = 1%of leaf areacoveredwith lesions; 3 = 5%; 5 = 10%; 7 = 25%; 9 = 50% or more of leaf area covered with lesions. zMeans followed by the sane letter are not significantly different at 5% level by Duncan's multiple range test. 69 Taylor Hort. was affected by the disease. This continued to be the trend even at 40 days after inoculation where, except for Cranberry Taylor Hort., there was a decrease in disease reaction for most cultivars. Few disease symptoms were observed on uninoculated trifoliolate leaves of I-84100 and Valley. Mean disease reactions of pods plus/minus standard error on plants at physiological maturity are shown in Table 2. Cranberry Taylor Hort. gave the highest disease rating, followed by Charlevoix and C-20. There was no significant difference in reactions of pods between I-84100 and Valley, as was the case for foliage reactions. Bacterial populations. Populations of XCP in reproductive tissues of various bean genotypes were studied at the flower bud, open flower, flat and bumpy pod stages of development. Due to differences in growth habits and differences in flowering dates of the different genotypes, sampling of parts was done at different number of days after planting, but at the same developmental stage. Except for pods, both surface and internal populations of XCP were determined. High numbers of blight bacteria were present on and in all reproductive tissues assayed. Data on surface populations of XCP on flower buds and open flowers are shown in Table 3. On Black Magic and C-20, surface populations were not significantly different from those on I-84100 and 70 Table 2. Reactions of pods on plants of seven bean genotypes at physiological maturity to Xanthomonas campestris pv. phaseoli (MI-17) under field conditions. Genotype Pod reactionz I-84100 1.8 1 0.17 d Valley 1.9 1_0.49 d Pinto UI-114 5.0 1 0.47 c Cranberry Taylor Hort. 8.0 1_ 0.47 a Charlevoix 7.3 1 0.29 ab Black Magic 5.5 1 0.33 c C - 20 6.9 1 0.49 b yEighteen-day-old bean plants were inoculated by gently spraying the foliage to run-off, without water sgaking, with a bacterial suspension containing ca 1.4x10 CFU/ml using a Knapsack sprayer. 2Means of four replications plus/minus standard error. Disease rating based on CIAT scale of 0-9; where 0 = immune, no symptoms; 1 = 1% of pod area covered with lesions; 3 = 5%; 5 = 10%; 7 = 25%; 9 = 50% or more of pod area covered with lesions. Means followed by the same letter are not significantly different at 5% level according to Duncan's multiple range test. 71 Valley. Surface populations on Valley, Pinto UI-114 and Cranberry Taylor Hort. were quite similar during the flower bud and the open flower stages; there were non-significant decreases in surface populations of XCP between the two stages on I-84100 and Black Magic. On C-20, surface XCP populations were higher on open flowers than on flower buds. However, the increase was not significant (Table 3). Some of the genotypes which showed lower susceptibility to the XCP isolate used, also supported higher internal population levels of the pathogen. This indicated that, under the conditions of this experiment, blight bacteria also multiplied in the less susceptible cultivars almost at the same rate as in susceptible cultivars. In all genotypes, flower buds and open flowers were heavily infected with blight bacteria. Population levels of XCP at both flower bud and open flower stages of development were not significantly different (P = 0.05) for genotypes I- 84100, Valley and Black Magic. Internal levels of XCP were similar to those observed for surface populations in all genotypes (Tables 3 and 4). These observations indicate that pods developing from such flower buds and flowers may be systemically infected with blight bacteria. Oily translucent cream-colored lesions which were more evident during dry sunny weather were observed on blossom petals of susceptible genotypes. Such symptoms were more 72 Table 3. Surface populations of Xanthomonas campestris pv. maseoli (MI-17) on flower buds and open flowers of various bean genotypes grown in the field." Stage of CFU/gran fresh weight of tissue Genotype DevelopmentY Rep 1 Rep 2 Rep 3 Rep 4 Meanz 3 4 3 3 4 I-84100 1 2.7x10 7.3x10 2.7x10 5.3x10 2.092x10 de 2 1 2x104 3 8x102 0 0 6.0x103 4 595x103 e 3 3 4 4 4 wnq 1 9&m ram 2mm 2am 2mam ma 2 4.5x104 1.1x103 2.5x103 3.3x105 9.465x1o4 bcde . 5 6 5 5 5 Pinto 01-114 1 2.7x10 1.7x10 1.1x10 5.5x10 6.575x10 ab 2 3 1x105 7 3x104 5 0x105 4 2x105 3 257x105 abc 6 5 5 5 5 cranberry’ 1 2.3x10 5.8x10 5.3x10 2.9x10 9.250x10 a Taylor Hort. 2 3.2x105 2.3x105 1.6x105 2.3x105 2.35011105 abcd . 3 6 5 6 6 Charlevorx 1 1.4x10 2.7x10 8.5x10 2.0x10 1.387x10 abc 2 1 9x105 4 1x105 8 5x104 1 2x106 4 712x105 abc . 4 4 3 2 4 Black Magic 1 1.8x10 2.5x10 3.4x10 2.8x10 1.167x1o e 2 7 8x103 4 6x103 7.2x102 4 3x103 4 355x103 e c -20 1 4.1x103 6.211102 3.221104 1.311105 4.168x10‘ ede 2 5.121105 3.211103 4.1x102 1.3x104 1.316x105 cde ’Tor surface populations. 25 flower buds or open flowers were shakai for 30 minutes in 40 ml of 0.01 M phosphate buffer pH 7.2 containing 10 17M Magnesium sulfate and 0.01% tween 20. After ten-fold serial dilutions in the sane buffer, 0.1 ml portions of washates were plated on M—SSM and plates incubated at 27 1 1C. Recipe for M-SS‘I median is described in the text (Chapter 1) . YStage of developnent: 1 -= flower buds; 2 = open flowers. zMeans followed by the same letter are not significantly different at 5% level by Duncan's nultiple range test. Data were log transformed before analysis. 73 Table 4. Internal populations of Xanthomonas campestris pv. phaseoli (MI-17) in flower buds and open flowers of various bean genotypes grown in the field.x CFU/ gram fresh weight of tissue Genotype Stage of develcpnenty Rep 1 Rep 2 Rep 3 Rep 4 Mean2 4 3 4 4 I-84100 1 0. 0 6 . 6x10 2 . 7x10 7 . 91110 3 . 692x10 cde ‘ ' 3 2 3 2 7.5x10 0.0 0.0 1.2110 4.875x10 e 3 4 5 4 Valley 1 7 . 3x10 0 . 0 1 . 3x10 1 . 2:10 3 . 507x10 cde 4 2 5 4 2 4.51410 2.31410 0.0 1.8x10 5.631x10 de . S 5 4 4 5 Pinto UI-114 1 1.6x10 1.4x10 4.5x10 9.3x10 1.095x10 abcd 2 1.2):105 1.614105 3.3)(105 1.8x105 1.9751(105 abcd 6 5 5 6 6 (kanberry 1 2 . 8x10 6. 8x10 6 . 8x10 6 . 4x10 2. 640x10 a Taylor Hort. 5 5 5 5 5 2 2.6x10 1.2110 2.0x10 1.3x10 1.775x10 abcd . 4 6 5 6 5 Charlevorx l 2 . 5x10 1 . 3x10 3 . 2:10 1 . 2:10 7 . 112110 ab 2 6 . 914104 1 . 91th5 5 . 01(104 3 . 07:106 8. 272x105 abc . 4 4 3 2 4 Black Magic 1 2 . 2110 3 . 3x10 2. 7x10 2 . 8x10 1 . 449x10 bode 2 1.6x103 3.511105 5.911103 1.721104 7.0 1:103 bcde c-2o 1 1.9x103 5.811103 3.511103 8.6x104 2.4 x 104 bcde 2 3.221105 5.71903 2.711103 2.711104 8.885x104 abcd xFor internal populations. 25 flower buds or open flowers were surface sterilized in 2.62% NaOCl for three minutes. rinsed three times with sterile distilled water and canninuted in 30 m1 of 0.01 M phosphate buffer pH 7.2. containing 10 nM M9504 and 0.01% tween 20. using sterile mortars and pestles. Suspensions were allowed to sit for 15 minutes followed by ten-fold serial dilution plating on M—SSM and incubation of plates at 27 1 1 C for 2-5 days. Recipe for M-SSG median is described in the text (chapter one). YStage of developnent: 1 - flower butb. 2 a open flowers. 7Means followed by the sane letter are not significantly different at 5% level by Duncan's nultiple range test. Data were log transformed before analysis. 74 prevalent on heavily infected cultivars such as Cranberry Taylor Hort., Pinto UI-114 and Charlevoix; which also experienced severe blossom drop. Blight bacteria were readily isolated from such lesions on M-SSM. Blossom petals on resistant genotypes did not exhibit these lesions. Surface populations of XCP on flat and bumpy pods were measured by plating washates from six pods for each replication on M-SSM. Bacterial populations on pods of each genotype were expressed as CFU/cm2 (Table 5). Differences in population levels of XCP between genotypes and between stages of pod development within genotype were observed. Line I-84100 supported significantly (P = 0.05) lower populations of blight bacteria during the flat stage than bumpy stage of pod development; and was also significantly lower than all other genotypes except Valley. Cranberry Taylor Hort. and Charlevoix supported higher but non- significant levels of blight bacteria in the flat stages than in bumpy stages. Populations of XCP on Black Magic, C- 20 and Pinto UI-ll4 were almost the same for both stages of pod deveIOpment (Table 5). Epiphytic bacterial contaminants. Populations of epiphytic bacterial contaminants which were associated with bean reproductive tissues are shown in Tables 6 and 7. Levels of bacterial contaminants were almost stable on genotypes I-84100 and C-20 for both flower bud and open 75 Table 5. Surface populations of Xanthomonas campestris pv. maseOIi (MI-17) on flat and bmpy pods of various bean genotypes groom in the field.x Stage of CFU/on2 pod area Gemtype Developmenty Rep 1 Rep 2 Rep 3 Rep 4 Mean2 I-84100 1 4.2x10 4.0x10 2.3x102 3.811102 1.7341102 f 2 2.741105 7.614103 1.721103 1.611104 7.383x104 abc 3 2 2 3 Valley 1 5.3x10 1.8x10 5.6x10 2.13110 1.5154110 cf 2 3 2 2 2 1.8x10 2.7x10 5.4x10 5.2x10 9.850x10 def . 3 4 4 3 4 Pinto 01-114 1 . 7.1x10 1.13110 5.6x10 2.7x10 1.9202110 abc 2 4.221103 2.811103 1.711104 2.511103 6.62521103 bed 3 5 5 4 5 cranberry 1 8.9x10 2.3x10 2.8x10 6.0x10 1.447x10 a Taylor Hort. 5 4 4 4 4 2 1.8x10 2.2x10 4.4x10 1.3x10 6.475x10 ab . 5 3 5 3 5 cremation 1 1.54110 1.8x10 5.51110 9.9x10 1.7791110 ab 2 5 8x104 2.111103 3.6x103 6.811103 1.76324104 abcd . 4 3 2 3 3 Black Magic 1 2.8x10 1.0x10 4.22110 2.31110 7.930x10 cde 2 2.541104 2.184103 7.511102 6.21102 7 29211103 cde c - 20 1 7.611104 3.014103 1.144104 1.2::10‘1 2.550903 abc 2 3.814102 2.721103 3.311103 1.321103 1.92051103 cde xSix symptcmless pods per replicate at both flat and bunpy stages were shaken for 30 minutes in 60 ml of sterile 0.01 M phosphate buffer. pH 7.2: containing 10 nM magnesiun sulfate and 0.01% tween 20. on a horizontal shaker (75 x 1.5 inch strokes per minute) . is described in the text (chapter one). YStages of develcpnent: 1 a flat pods: 2 = bunpy pods. Washates were decimally diluted in the same buffer: aliquots of 0.1 ml were plated on M-SSM and incubated for 5 days at 27 _+_ 1 C. Recipe for M—SSM median zMeans followed by the sane letter are not significantly different at 5% level by Duncan’s nultiple range test. Data were log transformed before analysis. 76 flower stages. This high number of bacterial contaminants on these genotypes may have accounted for relatively lower levels of surface blight bacteria (Tables 3 and 6). There was a significant decrease in populations of bacterial epiphytes (P = 0.05) between flower bud and open flower stages on cultivar Valley. The same population trend was observed for cultivars Charlevoix and Black Magic (Table 6). On pods, levels of bacterial epiphytes also remained relatively stable between flat and bumpy stages of pod development. However, a significant decrease (P = 0.05) was observed on cultivar C-20. In general, differences in population levels of bacterial epiphytes on pods (expressed as CFU/cm2 area) between genotypes were not statistically significant. Fungal contaminants were rarely recovered from flower buds, open flowers, and pods during assay, despite the use of the non-selective medium, YCA. This might be related to the high numbers of bacteria which might have prevented fungal propagules from germination and development. Pathogenicity1tests. Randomly selected XCP-like colonies from plates of M-SSM proved to be virulent strains of the pathogen. Attempts were also made to determine if various bacterial contaminants isolated from bean reproductive tissues during this study were able to cause 77 Table 6. Oontparative population levels of bacterial contaminants fron flower buds and 0pm flower washates of various bean genotypes grown in the field.w Stage of CPU/gran fresh weight of tissue Genotype Developmentx Rep 1 Rep 2 Rep 3 Rep 4 Meany 1-84100 1 1.611106 1.6x106 1.611106 1.711107 1.625x106 ab 2 1.321106 9.221105 6.914105 5.6x105 8.675x105 b 6 6 6 6 6 Valley 1 2.41110 1.9x10 2.0x10 2.4x10 2.175x10 a 2 1.311106 5.6x105 6.5x105 9.4x105 8.6251(105 b . 6 5 5 5 5 Char1evoix 1 2.02110 4.61410 9.8x10 4.9x10 9.8251110 b 2 7.271104 2.6x105 2.021105 1.24105 1.6 21105 c . 5 5 6 5 5 Black Magic 1 7.1x10 6.1x10 1.1x10 6.0x10 7.55 x10 b 2 2.711105 2.321105 2.911105 3.614105 2.875x105 c c-2o 1 8.111105 1.711106 1.041106 3.154106 1.6521106 ab 2 6.3x105 3.971105 3.121106 6.114105 1.182x106 b Pinto 01-114 NTz NT NT NT NT Cranberry Taylor Hort. NT NT NT NT NT wTwenty five flower buds or open flowers were shaken for 30 minutes in 40 ml of 0.01 M phosphate buffer pH 7.2. containing 10 nM Magnesiun sulfate and 0.01% tween. on a horizontal shaker (75x1.5 inch strokes/minute). After ten-fold serial dilutions in the sane buffer. 0.1 ml portions of washates were plated on YCA (recipe is described in the text. chapter one) in triplicate and incubated at 27 1 l C. Colony counts were made after 5 days. xStage of developnent: 1 - flower buds. 2 a open flowers. YMeans followed by the sane letter are not significantly different at 5% level by [hmcan's nultiple range test. Data were log transformed before analysis. 'NT - not tested. 78 Table 7. Population levels of bacterial contaninants in washates from flat and bumpy pods of various bean genotypes 9mm in the field.x Stage of CPU/m2 pod area Genotype DevelopmentY Rep 1 Rep 2 Rep 3 Rep 4 Mean2 4 3 3 4 3 I-84100 1 1.12410 7.52410 6.7x10 .2290 9.300x10 abc 2 3.411104 6.0x103 8.3x103 .321103 .39021104 abc Valley 1 2.721103 5.121103 3.121103 .721103 .40021103 c 2 6.724.103 6.914103 3.614103 .51103 .675x103 bc . 4 4 4 3 4 pinto 01-114 1 1.0x10 1.52110 2.32110 .0x10 .275x10 abc 2 1.221104 4.421103 2.814104 .8x104 .560x104 abc 3 4 5 4 4 Cranberry 1 1.72110 2.02110 1.4x10 .3x10 .618x10 ab Taylor Hort. 5 4 3 4 4 2 1.12110 1.4x10 9.52110 .7x10 .263x10 a . 4 4 5 3 4 Charlevoix 1 4.62410 5.22110 2.42110 .5x10 .638x10 a 2 2.421104 5.614103 4.221103 .8x104 .54521104 abc . 4 4 4 4 4 Black Magic 1 5.22110 2.6x10 1.22110 .5x10 1252410 ab 2 1.924104 2.321103 6.821103 .6x103 92521103 bc c-2o 1 1.221104 4.6x104 3.521104 .721105 .57524104 8 2 2.921103 1.121104 8.0x102 .8x103 .875x103 c xSix symptanless pods per replicate at both flat and bumpy stages were shaken for 30 minutes in 60 ml of sterile 0.01 M phosphate buffer. pH 7.2: containing 10 11M Magnesiun sulfate and 0.01% tween 20. on a horizontal shaker (75 x 1.5 inch strokes/minute) . medimn is described in the text (chapter one). yStages of development: 1 a flat pods: 2 a bumpy pods. Recipe for YCA Washates were decimally diluted in the same buffer: aliquots of 0.1 ml were plated on YCA and incubated for 5 days at 27 1 1C. zMeans followed by the sane letter are not significantly different (p a 0.05) by Duncan's nultiple range test. Data were log transformed before analysis. 79 diseases on bean plants grown in the greenhouse. None of these contaminants were pathogenic on bean. However, no efforts were made to reisolate these organisms from the inoculated tissue to check whether they survived and colonized the inoculated tissue. Seed infection. Since results from earlier studies with artificial suture inoculation had shown that seeds of resistant bean genotypes can be infected by blight bacteria, attempts were made to determine whether such seed infection could occur when inoculation occurs early during the seedling stage. To determine the presence of bacteria in bean seed, a total of 200 seeds randomly sampled from each replication were assayed for each genotype using M-SSM medium (Figure 1). Percent seed infection in various bean genotypes grown in the field and inoculated with XCP when 18 days old are-shown in Table 8. Cranberry Taylor Hort. exhibited the highest level of seed infection, followed by Pinto UI-114. For these two cultivars, some of the infected seeds were severely shrivelled. Charlevoix and C-20 were not significantly different (P = 0.05) in percent seed infection. Despite differences in foliage and pod reactions, I-84lOO and Black Magic did not differ significantly in percent seed infection. The lowest percentage of seed infection occurred in Valley (Table 8). 80 Figure 1. Recovery of Xanthomonas campestris pv. phaseoli from bean seed harvested from plants of line I-84100 grown in the field and spray inoculated when 18 days old. Medium was M—SSM and plates were incubated 3 days at 27 i 1C. 81 Table 8. Percent seed infection of various bean genotypes inoculated with Xantl'nncnas campestris pv. m under field conditionsx % seed infectiony Genotype Rep 1 Rep 2 Rep 3 Rep 4 MeanZ I-84100 4.0 4.0 2.0 5.5 3.88 Cd Valley 0.0 1.0 0.5 0.0 0.38 e Pinto UI-ll4 21.0 15.0 14.5 17.0 16.88 b Cranberry Taylor 28.5 43.0 32.5 33.5 34.38 a Hort. Charlevoix 3.0 4.5 6.5 5.5 4.88 c Black Magic 2.5 2.0 2.0 2.0 2.12 d C - 20 5.0 9.0 4.5 7.5 6.50 c xOne hundred seeds per replicate picked at random were surface— sterilized with 2.62% NaOCl for three minutes, rinsed three times with sterile glass distilled water and blotted dry on sterile paper towel. Seeds were placed hilum down on improved semi-selective medium (M-SSM) containing 1% bacto-agar and incubated at 27 1 1C for 2—5 days before evaluatim. For each replication, 10 seeds showing no evidence of infection were rechecked by a six—hour incubation (individually) in 4 ml of 0.01 M phosphate buffer pH 7.2, containing 10 11M Magnesium sulfate. The resulting suspension was spread on M—SSM and YCA and incubated as described previously. yValues are means of two experiments. A total of 200 seeds were used per replicate. 2'Means followed by the same letter are not significantly different at 5% level by Duncan's multiple range test. Data were arcsine transformed before analysis. 82 Most of the seeds which were internally infected with XCP in resistant genotypes (I-84100 and Valley) were symptomless. This was also true for Black Magic, where infected seeds exhibited no discoloration and shrivelling of the seed coats. High levels of seed infection paralleled high numbers of internal XCP in flower buds and open flowers for cultivars Cranberry Taylor Hort., Pinto UI-ll4 and Charlevoix; while lower seed infection in Black Magic, C-20, Valley and I-84100 was consistent with the lower levels of XCP on and in reproductive tissues of these genotypes. DISCUSSION The XCP pathogen was recovered in large numbers from flower buds, blossoms, pods, and seeds of all genotypes inoculated at the seedling stage (18 days old). Previous work (9, 81) indicated that XCP is capable of colonizing young tissues as they develop from the apical meristem; thus establishing a population gradient in the bean canopy. At least 5x106 bacterial cells per cm2 leaf tissue are needed for symptom development, a requirement which creates a latent period after tissues are initially colonized (81). Development of symptoms on tissues of resistant cultivars took longer than symptoms on tissues of susceptible cultivars. Thus symptoms on I-84100 and Valley were delayed for 7 days as compared to susceptible genotypes. It was also noted that reproductive tissues of susceptible cultivars supported higher populations of XCP than those of resistant genotypes. Except for a few cases, flower buds supported smaller XCP populations than blossoms. This indicates that a population build-up occurs over time (Tables 3 and 4). Similar trends were observed on pods,‘ especially for cultivars 1-84100 and Cranberry Taylor Hort. (Table 5). '83 84 Populations of the pathogen present on and in flower buds and blossoms obviously play a role in early colonization of developing pods. Other studies have presented evidence that there is systemic colonization of pods by XCP (9, 81, 83, 84). However, it is not well known whether the low populations of blight bacteria on and in reproductive tissues of resistant genotypes is due to reduced systemic movement or to other internal and external factors related to the host and the environment. Both external and internal factors have been reported to affect multiplication of blight bacteria on and in bean tissues. Weller and Saettler (81) indicated that bud colonization is an important mechanism of spread for XCP into the leaf canopy. They further suggested that, although multiplication of XCP occurs in buds, constant recolonization by blight bacteria in rain and dew water which runs down the petioles and into axils appear more important in maintaining bud infection. In some cases, resistant genotypes supported higher populations of bacterial contaminants than susceptible genotypes (Tables 6 and 7). The high number of bacterial contaminants may have accounted for low XCP populations on these genotypes through competition for nutrients and other mechanisms discussed by Blakeman and Brodie (5) and Leben (69). Population levels of resident bacteria are greatly influenced by climatic factors; thus decline in numbers of 85 the pathogen could be attributed to several factors including wash-off by rain. Nutritional differences and the amount of exudates from resistant and susceptible bean genotypes could further account for differential support of XCP populations. Fungal contaminants were rarely recovered from flower buds, blossoms, and pods. This might be related to the high numbers of bacteria which might have prevented fungal propagules from germination and development. Failure of germination of fungal propagules on plant surfaces due to high numbers of bacterial cells in their surroundings have been reported. Blakeman and Brodie (5) postulated that such failures result from nutrient uptake into the polysaccharide sheath of adjucent bacteria, leading to an increased steep diffusion gradient of nutrients from within to outside the fungal propagule. The results of this study also indicate that under field conditions, XCP can infect seeds of both susceptible and resistant bean genotypes when the pathogen is introduced into young seedlings. This is contrary to previous findings by Cafati and Saettler (9), who recovered XCP only from seed samples harvested from symptomless pods of field grown plants inoculated by spraying at the small-flat stage of plant development. High levels of percent seed infection corresponded with high numbers of internal blight bacteria 86 in flower buds and blossoms for cultivars Cranberry Taylor Hort., Pinto UI-114, and Charlevoix. Low percent seed infection in Black Magic, C-20, Valley and I-84100 agreed with low levels of XCP on and in reproductive tissues of these genotypes (Tables 3, 4, 5, and 8). These findings suggest that blight bacteria in flower buds and blossoms play a role in systemic infection of pods, resulting in infected seeds. A relationship between foliage susceptibility and internal seed infection has also been reported (65) for some bean cultivars. However, differential reaction of foliage and pods to XCP is also known to occur (77). Most of the seeds in resistant genotypes which were internally infected with XCP had no symptoms of internal infection. Similar observations were made for Black Magic. Symptomless infected seed have been reported to be most important in the epidemiology of bean blight because germination and seedling growth of such seeds are most similar to that of uninfected seed. This is in contrast to severely infected seed which germinate poorly or produce deformed seedlings, thus reducing the effectiveness as primary inocula sources (80). Infection of seed of resistant bean genotypes is a major concern in the bean industry because seed transmission is more apt to establish the pathogen in a new locality. Moreover, quarantine regulations generally are less strict 87 for seed than infected or infested plants or soil. Seed transmission also introduces the pathogen randomly throughout the field and provides numerous foci for primary infection. Such inoculum is more effective than that spreading from the margins of the field (3). Furthermore, symptomless seed is difficult to detect due to the low frequency in seed lots and low populations of the pathogen within the seed (80). Since resistant bean genotypes can become infected with XCP under natural conditions, studies are needed to determine the ability of such seeds to serve as.a source of primary inoculum under field conditions. To insure freedom from seed infection, several methods for detection of XCP in seed lots exchanged between various bean growing areas should be used. CHAPTER THREE ISOLATION OF EPIPHYTIC BACTERIAL ANTAGONISTS FROM VARIOUS BEAN GENOTYPES AND THEIR POTENTIAL FOR CONTROL OF XANTHOMONAS CAMPESTRIS PV. PHASEOLI 88 INTRODUCTION Aerial surfaces of plants are very selective habitats for microbial growth. Ample evidence for the occurrence of epiphytic microflora have been given (5, 22, 27, 34, 36, 37, 48, 51). Some of these organisms are deposits of airflora or result from the activities of insects or other animals. However, some of the flora is capable of multiplying on the surface of the plants (4, 55). Leben (37) reported that bacteria are believed to be the first residents of seedlings and a few leaves on the evergreen types of plants in many plant ecological systems. A higher relative moisture level is necessary for growth of these microorganisms. He also suggested that, under favorable conditions bacterial residents on one plant may be capable of migrating to other species as well; migration was dependent on delicate balance of many factors. Because rapid changes can occur in the physical environment around shoot surfaces, the size of bacterial populations is highly variable. However, the residual surviving inoculum is normally sufficient to give rise to large cell numbers when favorable conditions for multiplication occur. Such conditions usually coincide with 89 90 those required for infection of many plant pathogens (5). The in-depth exploration of foliage microflora on various crops have stimulated an interest in managing microflora for disease control. There is an extensive literature on antagonistic interactions between epiphytes and plant pathogens in relation to development of possible biological control methods (31, 33, 55, 67). Successful strategies for biological control of crown gall (Agrobacterium tumefaciens) (31) and economic considerations of disease control costs also encourage the use of microorganisms to control foliar diseases (20, 69). Bacterial antagonists isolated from resident microflora populations would seem to be good candidates for controlling foliar diseases. Blakeman and Brodie (5, 69) and Leben (34) have discussed several basic mechanisms of antagonism, which include: direct parasitism, production of extracellular antibiotics and other substances, competition for sites on the host, and stimulation of host defenses. The authors suggest that knowledge of these mechanisms is essential for proper employment of biological control, much like the structure-activity studies which are used in fungicide research. Bean common blight is generally managed through preventive measures such as use of pathogen-free seed and tolerant varieties (6). Interactions between Xanthomonas campestris pv. phaseoli (XCP) and other resident bacterial 91 microflora on bean plants have not received much attention. Thus, a clear understanding of such interactions is needed to determine the potential of epiphytic bacterial flora for disease control (6, 42). The occurrence of epiphytic bacterial contaminants in large numbers on some bean genotypes used in previous studies (Chapter 2) and other successful reports on biological control of some diseases (25, 31, 73) prompted the present investigations. This study reports on the isolation of epiphytic bacterial antagonists from various bean genotypes and their potential for control of XCP. MATERIALS AND METHODS Bacterial isolates. A total of 22 resident bacterial isolates obtained from different reproductive tissues in the previous studies (Chapter 2) were screened for in 31:59 ability to inhibit XCP (MI-17). Washates were obtained from flower buds, blossoms, flat and bumpy pods as described in Chapter Two. The washates were then decimally diluted and 0.1 ml aliquots from each dilution plated on YCA (yeast extract 10.09; Calcium carbonate 2.59; bacto-agar 15.0g, and glass distilled water 1000 ml) and incubated at 27 i 1C for 5 days. Single bacterial colonies were classified as presumed contaminants based on colony morphology and color as described by Pontius (53) and further purified by a series of transfers on YCA. If more than one isolate with identical colony characteristics and color was obtained in a given sample, or resembled any of the already purified ones, it was discarded. All bacterial contaminants were maintained in phosphate buffer (0.01 M, pH 7.2)/glycerol (40%) mixture at 5-7 C. Prior to use, bacteria were streaked on NBGY (Nutrient broth glucose agar containing nutrient broth [Difco] 8.09; glucose 5.09; bacto-agar 15.0 9, glass distilled water 1000 ml) and incubated at 27 i 1 C. 92 93 Pathogenicity tests. All bacterial contaminants were tested for pathogenicity on susceptible Charlevoix bean plants. Bean seedlings were grown in the greenhouse and inoculated when possessing two to three trifoliolate leaves (21 days old). Bacterial suspensions were prepared from 24- 48 hour-old NBGY cultures using 0.01 M phosphate buffer at pH 7.2. The suspensions were adjusted to an optical density of 0.1 at 620 mm (ca 1.7 - 3.9 x 107 CFU/ml). Twenty one— day-old bean plants were inoculated with each bacterial contaminant by infiltrating the cell suspension into trifoliolate leaves using a sterile 3 cc disposable hypodermic syringe. Infiltration was done abaxially by pressing the end of the syringe against the finger-supported leaflets and slowly introducing in bacterial suspension. Four spots, each of about 10 mm diameter, were infiltrated into each leaflet and four plants were used per isolate. An isolate of XCP (MI-17) and a sterile phosphate buffer were included as positive and negative controls, respectively. Inoculated plants were maintained in the greenhouse with temperatures ranging from 25-30 C and were observed for 16 days on daily basis for symptom development. After rating, plants were kept for sufficient time and observed further for any development of symptoms in heterologous combinations. In vitro screening for antagonism. Experiments were conducted to determine whether the bacterial epiphytes 94 isolated could influence the development of common blight, disease. Antagonistic activity of bacterial isolates was tested by the overlay method as described by Pontius (53). NBGA was used for both upper soft layer (nutrient broth 8.0g; glucose 5.0 g; bacto-agar 10.0 g; glass distilled water 1000 ml) and the lower layer which contained 15.0 g of bacto-agar. Inoculum for overlays was obtained by washing 24-48 hour-old NBGA cultures grown at 27 i 1C with phosphate buffer (0.01 M, pH 7.2). The suspensions were adjusted to an optical density of 0.25 at 620 nm. Decimal milliliter portions of the pathogen suspension was mixed into 4.0 ml of soft NBGA and vortexed for 5 to 10 seconds at medium speed to obtain a uniform distribution of the pathogen. The mixture was immediately poured on top of a 10 ml base layer and plates were left in the laminar flow transfer chamber for half an hour to dry. After drying, four droplets, each of 2 microliter volume of bacterial contaminants were spotted onto pathogen inoculated plates. Control plates were spotted with sterile phosphate buffer. Plates were carefully incubated at 27 i 1C and after 48-72 hours, they were examined for presence of inhibition zones and the size of zones recorded. Two experiments were conducted for each isolate and experiments were replicated three times. 95 In vivo screening for antagonism. Bacterial contaminants which gave zones of inhibition during the in 21:39 assay were tested further for ability to inhibit XCP in 2139 under greenhouse conditions. Bean plants of the Charlevoix variety were grown in the greenhouse in 10 cm diameter clay pots (one plant per pot) containing a 3:1 mixture of soil and vermiculite, respectively. Plants were maintained at temperatures ranging from 20-31C and were watered twice daily with tap water. Inoculum was prepared from 24-hour old cultures grown on NBGA medium. Bacterial cells were washed from the medium with sterile phosphate buffer and the resulting suspensions were adjusted to an optical density of 0.25 at 620 nm. Inoculum of XCP was prepared by washing cells from 24-hour old culture using the same buffer. Suspensions were adjusted to an optical density of 0.2 at 620 nm and further diluted to obtain the desired concentration of cells for any particular experiment (Ca 2 x 107 CFU/ml). Number of viable cells was always estimated by plate counts. Plants were spray-inoculated to run-off without water- soaking with candidate bacterial contaminants at four-day intervals starting when plants possessed two fully expanded trifoliolate leaves. Control plants were spray-inoculated with sterile phosphate buffer. Four treatments were used and each treatment was replicated five times. Treatment 1 = control, sprayed with phosphate buffer alone, 2 = sprayed 96 once with antagonist 2 days before inoculation with common blight bacteria; 3 = sprayed twice at 4 days interval (6 and 2 days before inoculation), 4 = sprayed three times at 4 day intervals (10, 6 and 2 days before inoculation). Two days after the last treatment with bacterial contaminants, plants (when blossoming) were spray-inoculated with XCP to run-off without water-soaking, and observed for symptom development on a daily basis. Disease rating was done progressively on the top three trifoliolate leaves using a CIAT scale of 0—9 where: 0 = No symptoms observed; 1 = 1% of leaf area covered with lesions; 3 = 5%; 5 = 10%; 7 = 25% and 9 = 50% or more of leaf area covered with lesions. Two experiments were conducted for each epiphytic bacterial antagonist. Statistical analysis. Data were analysed using MSTAT program version 4.0 (Department of Crop and Soil Science, Michigan State University, East Lansing). Significant differences between treatment means were estimated using Duncan's multiple range test. RESULTS Pathogenicity tests. Naturally occurring resident bacteria were isolated from reproductive parts of various bean genotypes. Twenty-two epiphytic bacterial contaminants were isolated and tested individually for pathogenicity on susceptible Charlevoix bean plants. All isolates were non- pathogenic on bean (Table 1 and Figure 1). Inoculated plants were kept up to 25 days before being discarded. No efforts were_made to reisolate these organisms from the inoculated tissue to determine whether they were able to multiply despite lack of symptoms on the host. In vitro screening for antagonism. Antagonistic activity of resident bacterial isolates were individually tested for ability to inhibit growth of XCP in 31:59. Of all 22 epiphytic bacterial contaminants screened, only three (13.6%) inhibited the pathogen in 31559 (Table 2). Zones of inhibition in the lawns of XCP (MI-17) were observed around colonies of isolates 4, 8 and 13, with inhibition zones of 2.1, 6.5 and 6.3 mm, respectively (Figure 2). Inhibition zones were measured from the colony 97 98 Table l. Epiphytic bacterial contaminants isolated from bean plants grown in the 2112111.a No Color on YCA Source(Genotype) Tissue obtained Pathogenicityb from 1 White Black Magic Open flowers -C 2 Light pink Black Magic Open flowers - 3 Orange Black Magic Flower buds - 4 White Black Magic Open flowers - 5 Orange C - 20 Flower buds - 6 Yellow C - 20 Flower buds - 7 Yellow Cranberry Taylor Hort. Bumpy pods - 8 Yellow Cranberry Taylor Hort. Flat pods - 9 White Pinto UI-114 Flower buds - 10 Creamy white Charlevoix Open flowers - 11 Orange I-84100 Bumpy pods - 12 Creamy white 1-84100 Flat pods - 13 White Valley Flower buds - 14 Creamy white Valley Open flowers - 15 Light yellow I-84100 Open flowers - 16 Yellow I-84100 Open flowers - 17 White 1-84100 Flower buds - 18 Light yellow I-84100 Open flowers - 19 Yellow I-84100 Flat pods - 20 Pink C - 20 Flat pods - 21 Light pink Pinto UI-ll4 Bumpy pods - 22 Yellow Black Magic Bumpy pods — aSingle colonies of bacterial contaminants were selected based on colony morphology and color. They were purified by a series of transfers on YCA. bPathogenicity was tested by infiltrating leaflets of been plants with bacterial suspensions adjusted to an optical density of 0.1 at 620 (ca 1.7-3.9 x 107CFU/m1). plants were used. c(-) - non-pathogenic. negative host reaction. Twenty one-day-old green house grown Charlevoix 99 Figure l. Pathogenicity test of epiphytic bacterial contaminants isolated from reproductive tissues of beans grown in the field. (A) Pathogenic XCP (MI-l7). (B) Bacterial contaminant No. 22. Figure 2. Zones of Xanthomonas campestris pv. phaseoli inhibitions around colonies of bacterial antagonist No. 8. Bacterial contaminant No. 1 in the same lawn is without inhibition zones. NBGA was used for both basal layer and the overlay medium (containing 1% agar). 100 Figure 1 Figure 2 101 Table 2 . _I_r_1 vitro screening of epiphytic bacterial contaminants from bean. plants for ability to inhibit growth of Xanthomonas campestris PV- M111 (MI-17)a Contaminant number Inhibiticnb Size of inhibition zone (mm)° O BNBHHHHHHHHHHOQQGMDOONH H omflmtflwal-‘O I I OOOOOPPOPQPOPOQOPOEQOOO OOOOOOOOOQOOOOMOOOI—‘OOO aNutrient Broth Glucose agar (NBGA: nutrient broth (Difco) 8.0 9; glucose 5.0 g; bacto-agar 15.0 g; distilled water 1000 ml) was used for both base layer and the upper soft layer (containing 10.0 g bacbo—agar). The upper soft layer was seeded with 0.1 ml of go. pv. Mi suspension adjusted to 0.25 at 620 m, vortexed for 5 to 10 secondsatmediunspeedandpouredmto lOmlofbase layer. Plates were left to dry for 30 minutes and four droplets, each 2 microliter in size, of bacterial contaminants were spotted onto the plates in triplicate. Plates were incubated at 27 1 1C and inhibition zones measured after 48-72 hours. b(-) = no zone of inhibition present; (+) = zone of inhibition present. Zones were measured from the colony of antagonist to the edge of XCP growth. 0Means of two experiments, each replicated three times. 102 of the respective antagonist to the edge of XCP growth. Antagonists 8 and 13 were tentatively identified as fluorescent pseudomonad (probably Pseudomonas fluorescens) and Bacillus sp, respectively. Antagonist 8 was a Gram negative rod, aerobic, produced a fluorescent pigment on KMB and was not pathogenic on bean. Antagonist 13 was a Gram positive rod with endospores and aerobic. Another antagonist, No. 4, which also produced zones of inhibition in the XCP lawn, was not tentatively identified. It was Gram negative, rod, aerobic, non-fluorescing on KMB and produced white colonies on YCA. In vivo screening for antagonism. The three bacterial isolates exhibiting in vitro antagonism to XCP were tested further in vivo. Disease severity ratings associated with the three antagonists tested are shown in Tables 3 and 4. For antagonist 8, been plants treated three times with the organism had significantly lower disease ratings than other treatments. Generally, plants treated with antagonists had reduced rates of disease development. Similar observations were made for antagonists 4 and 13 (Table 4). Compared to the controls, antagonist treated plants had smaller lesions initially, but which enlarged slowly with time (Figure 3). On treated plants, appearance of initial symptoms was 103 Table 3. E vivo screening of epiphytic bacterial antagonists for abili? to control Xanthomonas campestris pv. @a_se_oli in beans Disease rating/ days after Inoculatioanz Experiment Antagonist Treatmentx number number 16 20 24 1 8 1 5.2a 7.2a 8.3a 2 4.1a 7.9a 8.5a 3 3.7a 7.0a 7.9a 4 1.0b 3.6b 4.7b 10 days 14 days 18 days 2 8 1 2.6a 7.7a 8.7a 2 2.1ab 5.5ab 7.1ab 3 1.1bc 5.3b 6.4ab 4 0.4c 2.7c 4.2b wEighteen-day~old green house grown Charlevoix plants were spray- inoculated to runroff, with suspension of bacterial antagonist # 8 in 0.01 Miphosphate buffer pH 7.2, optical density 0.25 at 620 nm. Control plants were spray inoculated.with.phosphate buffer. Plants were then inoculated 2 days after the last antagonist application with cormon blight bacteria (MI-17), ca. 2 x 10 CFU/ml, and observed for sympton developrent on daily basis. xTreatments: 1 = Control (P04 buffer alone); 2 = sprayed.once with antagonist 2 days before inoculation with comm blight bacteria; 3 = sprayed twice at 4 day interval (6 and 2 days before inoculation); 4 = sprayed three times at 4 day interval (10, 6 and 2 days before inoculation). yDisease ratings were assigned to the top three trifoliolate leaves usingaCIATscaleofO-9;where0=inmme,msynptons;1=1% of the leaf area covered with lesions; 3 = 5%; 5 = 10%; 7 = 25%; 9 = 50% or more of the leaf area covered with lesions. values are means of five replicates. onr each experiment, means within the same column followed by the same letter are not significantly different at 5% level according to Duncanfs multiple range test. 104 Table 4 . In vivo screening of epiphytic bacterial antagonists for abili to control Xanthononas campestris pv. m in Antagonist Treatment Disease ratingz / Days after inoculation number 12er 12 16 20 4 1 4.1 a 7.1 a 8.5 a 2 4.7 a 6.9 a 7.7 ab 3 3.3 a 4.3 b 6.7 bc 4 1.8 b 3.1 b 6.5 c 13 days 17 days 20 days 13 1 6.7 a 8.4 a 8.9 a 2 4.7 ab 7.3 ab 8.5 a 3 5.5 ab 7.8 ab 8.8 a 4 2.9 b 6.5 b 7.9 b inghteen—day—old green house grown Charlevoix plants were spray- inoculated to run-off, with bacterial suspensions of bacterial antagonists 4 and 13, respectively, in 0.01 M phosphate buffer (pH 7.2), optical density 0.25 at 620 rm. Control plants were spray inoculated with phosphate buffer. Plants were then inoculated 2 days after the last antagonist application with cannon blight bacteria (MI-17) ca. 2 x 10 and observed for sympton developnent on daily basis. - , yTreatment: l = Control (P04 buffer alone); 2 = sprayed once with antagonists 2 days before inoculation with cannon blight bacteria; 3 = sprayed twice at 4 day interval (6 and 2 days before inoculation); 4=sprayedthreetimes at4dayintervals (10, 6and2daysbefore inoculation). zDisease ratings were assigned to the top three trifoliolate leaves using a CIAT scale of 0-9; where 0=inmune, no symptoms; 1=1% of the leaf area covered with lesions; 3=5%; 5=10%; 7=25%; 9=50% or more of the leaf area covered with lesions. Values are means of two experiments, each replicated five times. For each antagonist, means within the same colunn followed by the same letter are not significantly different at 5% level by Duncan's multiple range test. 105 delayed a minimum of 2-3 days. In all cases, repeated treatment with antagonists was essential to produce statistically significant protection. Control plants which received phosphate buffer alone were affected by disease much more rapidly and to a greater extent than plants receiving antagonist applications. DISCUSSION Non-pathogenic bacterial epiphytes isolated from beans were tested for effect on the development of the common blight disease. These studies were based on the observation that numbers of bacterial contaminants on reproductive tissues of tolerant bean genotypes were often greater than those of the pathogen. In addition, other studies have emphasized the importance of considering the activities of other microorganisms commonly found on the host when investigating the infection and development of other diseases (34, 42, 53, 73). Of twenty-two potential antagonists screened for in yitrg activity, only three (13.6%) produced zones of inhibition against XCP. Strong zones of inhibition were detected with antagonists 8 and 13 (Figure 2). Antagonist 4 gave smaller zones of inhibition. Although Pontius (53) reported that microbial antagonism in yitrg may be affected by incubation temperature, only one incubation temperature (27 i 1C) was used in the present study. This temperature is the Optimum for growth of XCP. Thus any bacterial organism producing antagonist activity at this temperature might be a promising biological control agent. 106 107 In much of the research on biocontrol of foliar diseases, isolates showing in 31:59 antagonism in the laboratory frequently prove unsuccessful in 3139 (69). Thus the three isolates inhibiting antagonism in 31:59 to XCP were tested further in 2139. All three antagonists delayed symptom appearance a minimum of 2-3 days; the rate of disease progress was also reduced. For all antagonists, repeated treatments were essential to produce statistically significant protection compared to the control. The need for repeated treatment to produce significant reduction of disease may arise from the fact that these antagonists take time to multiply to higher numbers due to variations in environmental conditions. The mechanism by which these bacterial antagonists reduce disease development was not investigated. Possibilities include changes in the host physiology which in turn leads to host defense mechanisms. Other mechanisms include those discussed by Blakeman and Brodie (5), Leben (34) and Spurr (69); such as production of extracellular antibiotics and other substances and competition for nutrients on the host. Studies toward this direction are needed in order to fully understand the potential of these organisms for biological control of Xanthomonas bean blight. It is encouraging that results obtained in vivo correspond with in vitro data. These results indicate that phylloplane microflora from beans influence development of bean blight, 108 and, as such antagonists are promising potential biocontrol agents for the disease. However, field studies are needed to support the current observations. More studies are also needed to provide a clear knowledge on the most effective mode of application of these antagonists. LITERATURE CITED Alvarez, A.M. and K. Lou. 1985. Rapid identification of Xanthomonas campestris pv. campestris by ELISA. Plant Disease 69: 1082-1086. Andersen, A.L. 1985. Common names for bean (Phaseolus vulgaris L.) diseases. Plant Disease 69: 653. Baker, K.F. and S.H. Smith. 1966. Dynamics of seed transmission of plant pathogens. Annual Review of Phytopathology 4: 311-334. Barnes, L.w. 1986. The future of phytopathological diagnostics. Plant Disease 70: 180. Blakeman, J.P. and I.D.S. Brodie. 1976. Inhibitions of pathogens by epiphytic bacteria on aerial plant surfaces. Pp 529-557. In: Dickinson, C.H. and T.F. Preece (Eds.). Microbiology of aerial plant surfaces. Academic Press. London. New York. 669 pp. Brathwaite, C.w.D. and R.S. Dickey. 1970. Synergism between Pseudomonas caryophylli and a species of Corynebacterium. Phytopathology 60: 1046-1051. Brock, T.D. 1974. Biology of microorganisms. Prentice-Hall, Inc. Englewood Cliffs, New Jersey, 852 PP- Burr, T.J. 1984. Isolation of phytopathogenic bacteria from soil and plant material using selective media and enrichment techniques. pp. 11-12. In: Quacquarelli, A. and F.J. Casano (Eds.). Phytobacteriology and plant bacterial diseases of quarantine significance. Instituto Sperimentale per La Patologia Vegetale. Roma, Italy. 135 pp. Cafati, C.R. and A.w. Saettler. 1980. Transmission of Xanthomonas phaseoli in seed of resistant and susceptible Phaseolus genotypes. Phytopathology 70: 638-640. 109 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 110 Canteros de Echenique, 8.1., D. Zagory and R.E. Stall. 1985. A medium for the cultivation of the B-strain of Xanthomonas campestris pv. citri, cause of cancrosis B in Argentina and Uruguay. Plant Disease 69: 122-123. Chun, W.W.C. and A.M. Alvarez. 1983. A starch- methionine medium for isolation of Xanthomonas campestris pv. campestris from plant debris in soil. Plant Disease 67: 632-635. Civerolo, E.L., M. Sasser, C. Helkie, and D. Burbage. 1982. Selective medium for Xanthomonas campestris pv. pruni. Plant Disease 66: 39-43. Civerolo, E.L. 1984. Techniques for detection and identification of phytopathogenic bacteria. Pp 7-10. In: Quacquarelli, A. and F.J. Casano (Eds). Phytobacteriology and plant bacterial diseases of quarantine significance. Istituto Sperimentale per la Patologia Vegetale, Roma, Italy. 135 pp. Claflin, L.E., A.K. Vidaver, and M. Sasser. 1987. MXP, a semi-selective medium for Xanthomonas campestris pv. phaseoli. Phytopathology 77: 730-734. Commonwealth Mycological Institute. 1973. Distribution maps of plant diseases. Map No. 402, Edition 2. Commonwealth Mycological Institute, Kew Survey, England. Copeland, L.O. and D.C. Greenman. 1976. Seed certification in Michigan. MSU Cooperative Extension Service. Extension Bulletin E-1019. No. 112. 6pp. Cuppels, D. and A. Kelman. 1974. Evaluation of selective media for isolation of soft-rot bacteria from soil and plant tissue. Phytopathology 64: 468-475. Dhanvantari, B.N. 1981. Semi-selective media for detection and monitoring of some Xanthomonas campestris pathovars. Proceedings of the 5th International Conference on Plant Pathogenic Bacteria, Cali, Colombia: pp. 135-136. Fahy, P.C. and A.C. Hayward. 1983. Media and methods for isolation and diagnostic tests. Pp 337-378. In: Fahy, P.C. and G.J. Persley (Eds). Plant bacterial diseases, a diagnostic guide. Academic Press, Sydney, New York, London, Paris, Tokyo, Toronto. 393 pp. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 111 Fraser, R.S.S. 1985. Mechanisms of resistance to plant diseases. Martinus Nijhoff/Dr. w. Junk Publishers. Boston. Lancaster. Dordrecht. 462 pp. Fung, D.V.C. and R.D. Miller. 1973. Effects of dyes on bacterial growth. Applied Microbiology 25: 793-799. Gardner, M.w. 1924. A native weed host for bacterial blight of beans. Phytopathology 14: 340. Goto, M. 1971. The significance of vegetation for the survival of plant pathogenic bacteria. Pp 39-52. In: Maas Geesteranus, H.P. (Ed). Proceedings of the Third International Conference on Plant Pathogenic bacteria. Wageningen, 14-20 April. University of Toronto Press, Toronto, Buffalo. 365 pp. Hayward, A.C. 1983. Preliminary diagnosis of plant diseases caused by bacteria. Pp 1—12. In: Fahy, P.C. and G.J. Persley (Eds). Plant bacterial diseases, a diagnostic guide. Academic Press. Sydney, New York, London, Paris, London, Tokyo, Toronto. 393 pp. Ishimaru, C.A. 1985. Herbicolins O, I and 2C, novel beta-lactam antibiotics produced by Erwinia herbicola strains C-9-1. Ph.D Thesis, Michigan State University, East Lansing, 108 pp. Kado, C.I. and M.G. Heskett. 1970. Selective media for isolation of Agrobacterium, Corynebacterium, Erwinia, Pseudomonas, and Xanthomonas. Phytopathology 60: 969-976. Kaneda, T. 1986. Seasonal population changes and characterization of ice-nucleating bacteria in farm fields of central Alberta. Applied and Environmental Microbiology 52: 173-178. Katznelson, H. 1950. The detection of internally- borne bacterial pathogens of beans by a rapid phage plaque count technique. Science 112: 263-266. Katznelson, H. and M.D. Sutton. 1951. A rapid phage plaque count method for the detection of bacteria as applied to the demonstration of internally-borne seed infection. Journal of Bacteriology 61: 689-701. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 112 Katznelson, H., M.D. Sutton, and S.T. Vayley. 1954. The use of bacteriophage of Xanthomonas phaseoli in detecting infection in beans with observation of its growth and morphology. Canadian Journal of Microbiology 1: 22-29. Kerr, A.- 1980. Biological control of crown gall through production of Agrocin 84. Plant Disease 64: 25-30. Kohne, D.E. 1986. Application of DNA probe tests to the diagnosis of infectious diseases. American Clinics in Laboratory Medicine. November. pp. 20-29. Leben, C. 1964. Influence of bacteria isolated from healthy cucumber leaves on two leaf diseases of cucumber. Phytopathology 54: 405-408. Leben, C. 1965. Epiphytic microorganisms in relation to plant disease. Annual Review of Phytopathology 3: 209-231. Leben, C. 1974. Survival of plant pathogenic bacteria. Special Circ. 1. Ohio Agricultural Research Development Center, Wooster, OH. 21 pp. Leben, C. and G.C. Daft. 1966. Migration of bacteria on seedling plants. Canadian Journal of Microbiology 12: 1119-1123. Leben, C., G.C. Daft, J.D. Wilson and H.F. Winter. 1965. Field tests for disease control by an epiphytic bacterium. Phytopathology, 55: 1375-1376. Liew, K.W. and A.M. Alvarez, 1981. Biological and morphological characterization of Xanthomonas campestris bacteriophages. Phytopathology 71: 269-273. Liew, K.W. and A.M. Alvarez. 1981. Phage typing and lysotype distribution of Xanthomonas campestris. Phytopathology 71: 274-276. Little, T.M. and F.J. Hills. 1978. Agricultural experimentation, design and analysis. John Wiley and Sons. New York. Brisbane. Toronto. Singapore. 350 PP- Loper, J.E., T.V. Suslow, and M.N. Schroth. 1984. Lognormal distribution of bacterial pOpulations in the rhizosphere. Phytopathology 74: 1454-1460. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 113 Maino, A.L., M.N. Schroth and V.B. Vitanza. 1974. Synergy between Achromobacter sp. and Pseudomonas phaseolicola resulting in increased disease. Phytopathology 64: 159-276. Malin, E.M. and D.A. Roth. 1983. Indirect immunofluorescent staining for detection and identification of Xanthomonas campestris pv. phaseoli in naturally infected bean seed. Plant Disease 67: 645-647. McGuire, R.G., J.B. Jones and M. Sasser. 1986. Tween media for the semi-selective isolation of Xanthomonas campestris pv. vesicatoria. Plant Disease 70: 887-891. Meynell, G.C. and E. Meynell. 1965. Theory and practice in experimental bacteriology. Cambridge University Press, London. 287 pp. Miller, T.D. and M.N. Schroth. 1972. Monitoring the epiphytic population of Erwinia amylovora on pear with a selective medium. Phytopathology 62: 1175-1182. Mulcahy, L. 1986. DNA probe, an overview. American Clinics in Laboratory Medicine. November. pp. 14-19. Mundt, J.O., J.H. Coggin, Jr. and L.F. Johnson. 1962. Growth of Streptococcus faecalis var. liquefasciens on plants. Applied Microbiology 10: 552-555. Noel, J.K. 1986. DNA probe vs monoclonal antibody. American Clinics in Laboratory Medicine. November. pp. 10-13. O'Neill, R. and C. Logan. 1975. A comparison of various selective isolation media for their efficiency in the diagnosis and enumeration of soft rot coliform bacteria. Journal of Applied Bacteriology 39: 139- 146. Perombelon, M.C.M. 1981. Ecology of erwinias on plant surfaces. Pp 411-431. 19: Blakeman, J.P. (Ed). Microbial ecology of the phylloplane. Academic Press. London. New York. Toronto. 502 pp. Persley, G.J. 1983. Bacteriophages for the identification of plant pathogenic bacteria. Pp 259- 273. In: Fahy, P.C. and G.J. Persley (Eds). Plant bacterial diseases, a diagnostic guide. Academic Press. Sydney. New York. London. Tokyo. Toronto. 393 pp. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 114 Pontius, L.T. 1983. In vitro and in vivo inhibition of bacterial and fungal pathogens of beans by bacterial antagonists. M.S. Thesis, Michigan State University, East Lansing. 89 pp. Randhawa, P.S. and N.W. Schaad. 1984. Selective isolation of Xanthomonas campestris pv. campestris from crucifer seeds. Phytopathology 74: 268-272. Reinecke, P. 1981. Antagonism and biological control on aerial surfaces of the gramineae. Pp 382-395. In: Blakeman, J.P. (Ed.). Microbial ecology of the phylloplane. Academic Press. London. New York. Toronto. 502 pp. Saettler, A.W., C.R. Cafati and D.M. Weller. 1986. Non-overwintering of Xanthomonas bean blight bacteria in Michigan. Plant Disease 70: 285-287. Schaad, N.W. 1979. Serological identification of plant pathogenic bacteria. Annual Review of Phytopathology 17: 123-147. Schaad, N.W. 1982. Detection of seed-borne bacterial plant pathogens. Plant Disease 66: 885-890. Schaad, N.W. and W.C. White. 1974. A selective medium for soil isolation and enumeration of Xanthomonas campestris. Phytopathology 64: 876-880. Schaad, N.W. and R.L. Forster. 1985. A semi-selective agar medium for isolating Xanthomonas campestris pv. translucens from wheat seeds. Phytopathology 75: 260- 263. Schnathorst, W.C. 1954. Bacteria and fungi in seeds and plants of certified bean varieties. Phytopathology 44: 588-592. Schneider, R.W. and R.G. Grogan. 1977. Bacterial speck of tomato; source of inoculUm and establishment of a resident population. Phytopathology 67: 388-394. Schuster, M.L. and D.P. Coyne. 1974. Survival mechanisms of phytopathogenic bacteria. Annual Review of Phytopathology 12: 199-221. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 115 Schuster, M.L. and D.P. Coyne. 1977. Survival of plant parasitic bacteria of plants grown in tropics with emphasis on beans (Phaseolus vulgaris L.). Fitopath Brasileira 2: 117-130. Schuster, M.L., D.P. Coyne, D.S. Nuland, and C.C. Smith. 1979. Transmission of Xanthomonas phaseoli and other bacterial species or varieties in seeds of tolerant bean (Phaseolus vulgaris) cultivars. Plant Disease Reporter 63: 955-959. Sheppard, J.W., D. Roth, and A.W. Saettler. (Unpublished). Detection of Xanthomonas campestris pv. phaseoli. Sinha, S. 1965. Microbiological complex of the phyllosphere and disease control. Indian Phytopathology 18: 1-20. Sly, L. I. 1983. Preservation of microbial cultures. Pp. 275-298. In: Fahy, P.C. and G.J. Persley (Eds). Plant bacterial diseases, a diagnostic guide. Academic Press. Sydney. New York. London. Tokyo. Toronto. 393 pp. Spurr, H.W. Jr. 1981. Experiments on foliar disease control using bacterial antagonists. Pp 368-381. In: Blakeman, J.P. (Ed). Microbial ecology of hte phylloplane. Academic Press. London. New York. Toronto. Sydney. 502 pp. Starr, M.P. 1945. The nutrition of phytopathogenic bacteria. Minimal nutritive requirements of the genus Xanthomonas. Journal of Bacteriology 51: 131- 143. Sutton, M.D. and V.R. Wallen. 1967. Phage types of Xanthomonas phaseoli isolated from beans. Canadian Journal of Botany 45: 267-280. Sutton, M.D. and V.R. Wallen. 1970. Epidemiological and ecological relationship of Xanthomonas phaseoli and g; phaseoli var. fuscans on bean in Southern Ontario, 1961-1968. Canadian Journal of Botany 48: 1329-1334. Teliz-Ortiz, M. and W.H. Burkholder. 1960. A strain of Pseudomonas fluorescens antagonistic to Pseudomonas phaseolicola and other bacterial plant pathogens. Phytopathology 50: 119-123. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 116 Thaveechai, N. and N.W. Schaad. 1986. Serological and electrophoretic analysis of a membrane protein extract of Xanthomonas campestris pv. campestris from Thailand. Phytopathology 76: 136-147. Trujillo, G.E. 1979. The use of serology and semi- selective media as aids in the detection of Xanthomonas bean blight bacteria. Ph.D. Thesis, Michigan State University, East Lansing. 132 pp. Trujillo, G.E. and A.W. Saettler. 1980. A liquid semi-selective medium for Xanthomonas phaseoli and Xanthomonas phaseoli var. fuscans. Research Report 411. MSU. Agric. Experiment Station, East Lansing, 7 PP- Valladares-Sanchez, N.E., D.P. Coyne, and M.L. Schuster. 1979. Differential reaction of leaves and pods of Phaseolus germplasm to strains of Xanthomonas phaseoli and transgressive segregation for tolerance from crosses of susceptible germplasm. Journal of American Society of Horticultural Science 104: 648- 654. Wallen, V.R. and M.D. Sutton. 1964. Xanthomonas phaseoli var. fuscans (Burkh) Starr and Burkh. on field bean in Ontario. Canadian Journal of Botany 43: 437-446. Webster, D.M., J.D. Atkin and J.E. Cross. 1983. Bacterial blights of snap beans and their control. Plant Disease 67: 935-940. Weller,-D.M. and A.W. Saettler. 1980. Evaluation of seed-borne Xanthomonas phaseoli and X. phaseoli var. fuscans as primary inocula in bean blights. Phytopathology 70: 148-152. Weller, D.M. and A.W. Saettler. 1980. Colonization and distribution of Xanthomonas phaseoli and g; phaseoli var. fuscans in field-grown navy beans. Phytopathology 70: 500-506. Yoshii, K. 1980. Common and fuscous blight. Pp 155- 172. In: Schwartz, H.F. and G.E. Galvez (Eds). Bean production problems. CIAT, Cali. Colombia. 424 pp. Zaumeyer, W.J. 1930. The bacterial blight of beans caused by Bacterium phaseoli. USDA Technical Bulletin No. 186. 36 pp. 117 84. Zaumeyer, W.J. and H.R. Thomas. 1957. A monographic study of bean diseases and methods of their control. USDA Technical Bulletin 868. 255 pp. ‘1111111111“