The microbiological basis of a soil suppressive to fusarium wilt of watermelon

MISSING IMAGE

Material Information

Title:
The microbiological basis of a soil suppressive to fusarium wilt of watermelon
Physical Description:
Book
Language:
English
Creator:
Larkin, Robert Philip, 1956-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 27771933
oclc - 26498488
System ID:
AA00022842:00001

Full Text










THE MICROBIOLOGICAL BASIS OF A SOIL SUPPRESSIVE
TO FUSARIUM WILT OF WATERMELON

















By

ROBERT PHILIP LARKIN















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1990.














ACKNOWLEDGEMENTS


I would like to express my sincere appreciation to Dr. Don Hopkins for his confidence, guidance, and continued support through all stages of this research and for providing me the opportunity to pursue this degree in plant pathology.

I would also like to thank Dr. Frank Martin for his guidance, counsel, and commitment throughout this work. In addition, I would like to thank Dr. Dave Mitchell for his expertise, counsel, and encouragement when I needed it most. I also extend thanks to committee members Dr. Corby Kistler and Dr. David Sylvia for their insightful input and cooperation. Additional thanks go to Charles R. Semer IV for his expertise and assistance on various technical aspects of this work. I also thank the Richard C. Storkan Foundation for providing partial support for this work.

Special thanks are extended to my friends in Gainesville for their friendship, warmth, and comraderie which really helped me get through the rough times.

Most importantly, I want to thank my parents for their confidence, support, and patience throughout this long and winding road I have taken. Although they always wanted me to be a doctor, I don't think this is quite what they had imagined.














TABLE OF CONTENTS



ACKNOWLEDGEMENTS............................................. I

LIST OF TABLES ...................................................vi

LIST OF FIGURES .................................................ix

ABSTRACT....................................................... A

CHAPTER 1 INTRODUCTION.........................................1

CHAPTER 2 VEGETATIVE COMPATIBILITY WITHIN FUSARIUM OXYSPORUM F.
SP. NIVEUM AND ITS RELATIONSHIP TO VIRULENCE,
AGGRESSIVENESS, AND RACE ..................................9
Introduction................................................. 9
Materials and Methods ........................................ 12
Isolates of Fusarium oxvs~orum. ........................... 12
Recovery of Nitrate-Nonutilizing Mutants...................... 13
Vegetative Compatibility Tests ..............................13
Pathogenicity Tests and Race Determinations ..................14
Results .................................................... 15
Vegetative Compatibility Tests ............................. 15
Pathogenicity Tests ..................................... 16
Race Determination ..................................... 16
Relationship of Vegetative Compatibility Group to Aggressiveness
and Race ....................................... 17
Discussion ................................................. 18

CHAPTER 3 ECOLOGY OF FUSARIUM OXYSPORUM F. SP. NIVEUM IN SOILS
SUPPRESSIVE AND CONDUCIVE TO FUSARIUM WILT OF
WATERMELON .............................................. 29
Introduction................................................29
Materials and Methods ........................................ 30
Soils Used ...................................30
Infestation of Soils with Fusarium ontslorum f. sp. niveum and
Assay of Fusarium Wilt ............................. 32
Production and Characterization of Orange Mutant Pathogen
Strains ......................................... 33



III








Survival and Population Dynamics of Fusarium oxvsporum in Field
Soils ........................................... 35
Chlamydlospore Germination .............................. 36
Root Colonization by Fusarium oxvsporum.................... 37
Soil and Root Microorganism Populations .....................38
Selective Elimination of Microorganisms by Microwave Irradiation . 39 Statistical Analyses......................................39
Results .................................................... 40
Production and Characterization of Orange Mutant Pathogen
Strains ......................................... 40
Survival and Population Dynamics of Fusarium oxvsporum in Field
Soils ........................................... 40
Chlamydospore Germination .............................. 41
Root Colonization by Fusarium oxvs~orum.................... 42
Soil and Root Microorganism Populations .....................43
Selective Elimination of Microorganisms by Microwave Irradiation . 45
Discussion ................................................. 47

CHAPTER 4 THE EFFECT OF SUCCESSIVE WATERMELON PLANTINGS ON
FUSARIUM OXYSPORUM AND OTHER MICROORGANISMS IN SOILS SUPPRESSIVE AND CONDUCIVE TO FUSARIUM WILT OF
WATERMELON .............................................. 71
Introduction...... .......................................... 71
Materials and Methods ........................................ 73
Soils Used ............................................ 73
Successive Plantings of Watermelon and Assay of Fusarium Wilt ... 74
Population Dynamics and Root Colonization of Fusarium
oxysporum ...................................... 74
Soil Microorganism Populations ............................ 75
Statistical Analyses......................................76
Results ..................................... ...............76
Population Dynamics of Fusarium oxysporum and Fusarium Wilt .. 76 Root Colonization by Fusarium oxvsporum.................... 79
Soil Microorganism Populations ............................ 81
Discussion ................................................. 82

CHAPTER 5 THE ROLES OF INDIGENOUS FUSARIUM OXYSPORUM AND
VARIOUS OTHER MICROORGANISMS IN A SOIL SUPPRESSIVE TO
FUSARIUM WILT OF WATERMELON ............................. 100
Introduction...............................................100
Materials and Methods ....................................... 101
Soil Infestation and Assay of Fusarium Wilt ...................101
Isolation of Potential Antagonists ...........................102
Screening Isolates in Microwave-Treated and Field Soils .........103 Characteristics of Successful Antagonists.................... 104
Induced Systemic Resistance............................. 105
Results ...............*........................106
Isolation and Screening of Potential Antagonists ...............106

iv








Characteristics of Successful Antagonists ................... 109
Induced Systemic Resistance ............................ 111
D iscussion ........................................... ..... 112

CHAPTER 6 SUMMARY AND CONCLUSIONS .......................... 134

APPENDIX A ADDITIONAL DATA FROM REPEATED EXPERIMENTS .......... 142 APPENDIX B ADDITIONAL SUCCESSIVE PLANTING EXPERIMENT .......... 155 LITERATURE CITED ............................................... 177

BIOGRAPHICAL SKETCH ........................................... 189





































v













LIST OF TABLES


Table Page


2-1 Isolates of Fusarium oxvsporum f. sp. niveum used in comparative
pathogenicity tests ........................................ 19

2-2 Wilt development in different watermelon cultivars planted in soil infested
with various isolates of Fusarium oxvsporum f. sp. niveum ........... 20

2-3 Average percent wilt in various watermelon cultivars caused by isolates of
Fusarium oxysporum f. sp. niveum in three vegetative compatibility
g ro u ps . . . . . . . . . . .. . . . . . . . . . . . . . 2 1

2-4 Vegetative compatibility groups and characteristics established for
Fusarium oxysporum f. sp. niveum ............................ 22

3-1 Comparison of selected orange mutant isolates with wild-type parent
isolate (FG85-1) of Fusarium oxsporum f. sp. niveum for growth
characteristics, root colonization, and pathogenicity ................ 48

3-2 Surface colonization of roots of two different watermelon cultivars by
Fusarium oxvsporum in four soils in relation to disease incidence and
populations of F. oxysporum f. sp. niveum ....................... 53

3-3 Internal colonization of 'Crimson Sweet' watermelon roots by Fusarium
oxysporum in two soils exposed to varying microwave treatments ..... 57

3-4 Microorganism populations on 'Crimson Sweet' watermelon roots in two
soils exposed to varying microwave treatments ................... 58

3-5 Rhizoplane organism populations in two soils exposed to varying
microwave treatments ...................................... 59

4-1 Surface colonization of roots of two watermelon cultivars by Fusarium
oxysporum in relation to soil populations and disease incidence in four
soils after four successive watermelon plantings .................. 89




vi








4-2 Internal colonization of 'Crimson Sweet' watermelon roots by Fusarium
oxvsporum in soils suppressive and conducive to Fusarium wilt following
four successive plantings .................................... 90

5-1 Potential antagonistic organisms isolated from watermelon roots in
suppressive and nonsuppressive soil and their ability to reduce disease in
infested, microwave-treated, suppressive soil ..................... 114

5-2 Isolates of Fusarium oxysporum most effective in reducing disease in
repeated antagonist screening tests ............................ 118

5-3 Effect of antagonist treatments in microwave-treated suppressive soil on
internal colonization of watermelon roots by Fusarium oxvsporum f. sp.
niveum and other F. oxysporum ............................... 119

5-4 Effect of antagonist treatments in conducive field soil on internal
colonization of watermelon roots by Fusarium oxysporum f. sp. niveum
and other F. oxvsporum ..................................... 120

5-5 Comparison of source and isolation conditions of isolates of Fusarium
oxvsporum effective in reducing disease in microwave-treated and field
so ils . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1

5-6 Development of Fusarium wilt in split-root watermelon plants with one half
of each root system exposed to the pathogen and the other half exposed
to conducive and suppressive soil treatments ..................... 122

A-1 Wilt development in different watermelon cultivars planted in soil infested
with various isolates of Fusarium oxvsporum f. sp. niveum (Results of
imported isolate test 1, which were combined with test 2 in Table 2-2).. 142

A-2 Wilt development in different watermelon cultivars planted in soil infested
with various isolates of Fusarium oxsporum f. sp. niveum (Results of
imported isolate test 2, which were combined with test 1 in Table 2-2).. 143

A-3 Wilt development in different watermelon cultivars planted in soil infested
with isolates of Fusarium oxvsporum f. sp. niveum from Florida (Florida
isolate test) .............................................. 144

A-4 Average percent wilt of various watermelon cultivars caused by isolates of
Fusarium oxvsporum f. sp. niveum in three vegetative compatibility
g ro ups .. ............. ... .. .. .. .. .. .. .. ..... .. ... ... . 145

A-5 Comparison of orange mutant isolates with wild-type parent isolate
(FG85-1) of Fusarium oxysporum f. sp. niveum for growth characteristics, root colonization, and pathogenicity (Initial tests on all orange mutant
iso lates) ........... ...... .. .. .. .. .. .. .... .... .. .. .. .. 146


vii








A-6 Microorganism populations on the roots of watermelon cultivars 'Crimson
Sweet' and 'Florida Giant' in four different soils (Combined data for
cultivars used in Figure 3-4) .................................. 147

A-7 Microorganism populations on 'Crimson Sweet' watermelon roots in two
soils exposed to varying microwave treatments (Repeat of experiment in
Table 3-4) ............................................... 148

A-8 Microorganism populations on 'Crimson Sweet' watermelon roots in two
soils exposed to varying microwave treatments and overall population
averages (Combined results of Table A-7 and Table 3-4) ............ 149

A-9 Surface colonization of roots of two watermelon cultivars by Fusarium
oxvsporum in relation to soil populations and disease incidence in four soils after four successive watermelon plantings (Early experiment similar
to that in Table 4-1) ........................................ 151

A-1 0 Effectiveness of isolates from various microorganism groups in reducing
disease in initial screening tests in microwave-treated soil (Summary of
results from all screening tests) ............................... 152






























viii













LIST OF FIGURES


Figure Page


3-1 Population dynamics of Fusarium oxysporum f. sp. niveum over time in
four soils under different moisture regimes ...................... 50

3-2 Chlamydospore germination of two different isolates of Fusarium
oxysporum f. sp. niveum in four soils after additions of glucose ....... 52

3-3 Population estimates of bacteria, actinomycetes, fluorescent
pseudomonads, other pseudomonads, and fungi in soil and on
watermelon roots in four different soils ......................... 55

3-4 Fusarium wilt development in three soils exposed to varying length
microwave exposures ...................................... 56

3-5 Rhizosphere and rhizoplane microorganism populations on 3-week-old
'Crimson Sweet' watermelon roots in two soils ................... 61

4-1 Population dynamics of Fusarium oxvsporum f. sp. niveum (as represented
by an orange-colored mutant pathogen) in four soils with successive
plantings of two different watermelon cultivars .................... 84

4-2 Population dynamics of indigenous Fusarium oxvsporum in four soils with
successive plantings of two different watermelon cultivars ........... 86

4-3 Fusarium wilt in four soils with successive plantings of two different
waterm elon cultivars ....................................... 88

4-4 Populations of bacteria, actinomycetes, fluorescent pseudomonads, other
pseudomonads, and fungi in four soils after four successive plantings of
watermelon cultivars 'Florida Giant' and 'Crimson Sweet'............. 92

5-1 Diagram of watermelon plant in split-root pot assembly ............. 113

5-2 Percent reduction of Fusarium wilt by various potential antagonistic
isolates of Fusarium oxvsporum, fluorescent pseudomonads, and other
bacteria in a typical screening trial in microwave-treated soil ......... 115


ix








5-3 Percent reduction of Fusarium wilt in conducive field soil screening tests
by various bacteria and nonpathogenic isolates of Fusarium oxysporum
collected from suppressive soil ............................... 117

A-1 Fusarium wilt in four soils with successive plantings of two different
watermelon cultivars (Repeat of test shown in Figure 4-3) ............ 154

B-1 Fusarium wilt development in four soils with successive plantings of four
different watermelon cultivars (Comparison by soil type for each cultivar;
average over all pathogen isolates)............................ 158

B-2 Fusarium wilt development in four soils with successive plantings of four
different watermelon cultivars (Comparison by cultivar for each soil;
average over all pathogen isolates)............................ 160

B-3 Fusarium wilt development in four soils with successive plantings of
watermelon cultivar 'Florida Giant' (susceptible to Fusarium wilt) for four
different isolates of Fusarium oxysporum f. sp. niveum. ............. 162

B-4 Fusarium wilt development in four soils with successive plantings of
watermelon cultivar 'Charleston Gray' (moderately resistant to Fusarium
wilt) for four different isolates of Fusarium oxysporum f. sp. niveum . 164

B-5 Fusarium wilt development in four soils with successive plantings of
watermelon cultivar 'Crimson Sweet' (moderately resistant to Fusarium wilt and inducer of soil suppressiveness) for four different isolates of Fusarium
oxysporum f. sp. niveum ......... ........................ .. 166

B-6 Fusarium wilt development in four soils with successive plantings of
watermelon cultivar 'Calhoun Gray' (highly resistant to Fusarium wilt) for
four different isolates of Fusarium oxysporum f. sp. niveum........... 168

B-7 Average populations of bacteria, actinomycetes, fluorescent
pseudomonads, other pseudomonads, and fungi in four soils after four
successive plantings of various watermelon cultivars............... 170

B-8 Populations of bacteria, actinomycetes, fluorescent pseudomonads, other
pseudomonads, and fungi in four soils after four successive plantings of
various watermelon cultivars (Comparison by soil type for each cultivar). 172

B-9 Populations of bacteria, actinomycetes, fluorescent pseudomonads, other
pseudomonads, and fungi in four soils after four successive plantings of
various watermelon cultivars (Comparison by cultivar for each soil type). 174

B-10 Population estimates of red-pigmented isolates of Fusarium oxysporum in
four soils after four successive plantings of various watermelon cultivars. 176



x













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

THE MICROBIOLOGICAL BASIS OF A SOIL SUPPRESSIVE TO FUSARIUM WILT OF WATERMELON

By

ROBERT PHILIP LARKIN

DECEMBER 1990

Chairman: Donald L. Hopkins Cochairman: Frank N. Martin Major Department: Plant Pathology

The nature of soil suppressiveness to Fusarium wilt of watermelon, caused by Fusarium oxvsporum f. sp. niveum, was investigated in a unique suppressive soil developed through monoculture of watermelon cultivar 'Crimson Sweet.' The objectives of this research were to evaluate various ecological characteristics of the suppressive soil in relation to the pathogen and identify the organism(s) and mechanism(s) responsible for suppressiveness.

In addition, a method for identifying and differentiating isolates of F. oxysporum f. sp. niveum using vegetative compatibility was developed. All isolates of F. oxvsporum f. sp. niveum belonged to one of three distinct vegetative compatibility groups (VCGs), and were incompatible with isolates not pathogenic on watermelon. Race 1 isolates were contained in 2 VCGs (0080 and 0081), while all race 2 isolates were contained in a third VCG (0082).




xi








Population dynamics, root colonization, and chlamydospore germination of F. oxysporum f. sp. niveum were monitored in relation to other microorganism populations, the incidence of Fusarium wilt, and the planting of watermelon cultivars in four different suppressive and conducive soils. An orange-colored mutant isolate of F. o xysporum f. sp. niveum, comparable to the wild-type pathogen in growth, pathogenicity, and root colonization, was used to differentiate the pathogen from indigenous populations of F. oxysporum in field soils. Suppressiveness was not associated with inhibition of chlamydospore germination or increased fungistasis. Suppressive soils maintained lower pathogen populations than conducive soils, even when planted to susceptible watermelon cultivars. Successive plantings of cultivar 'Crimson Sweet' caused changes in microflora populations, including increases in bacteria, actinomycetes, fluorescent pseudomonads, and nonpathogenic F. oxysporum. Root colonization by the pathogen and other indigenous F. oxvsporum was not consistently related to suppression.

Nearly 400 isolates of F. oxvsporum and miscellaneous bacteria, actinomycetes, and fungi were isolated from watermelon roots and tested for their ability to reduce disease in microwave-treated and field soils. Specific isolates of nonpathogenic F. oxvsporum were the only organisms consistently effective in reducing disease (35-75% reduction) in both soils. Isolate effectiveness was not related to the level of colonization of roots. The mechanism of suppression is still not clear, but may involve induced resistance.










xii














CHAPTER 1
INTRODUCTION


Fusarium oxvsporum (Schlect.) Snyd. & Hans. is an ubiquitous soil-inhabiting fungus, consisting of saprophytic as well as plant-pathogenic strains. Pathogens within this species are responsible for a severe and economically important vascular wilt disease on a staggeringly diverse group of plants and can be subdivided into numerous formae speciales, which are specialized pathogens of specific plant genera or other plant groups. These pathogens are capable of infecting the roots of susceptible host plants and extensively colonizing the xylem tissue, ultimately resulting in severe internal water stress producing wilt and eventually causing the death of the plant (Beckman, 1987; MacHardy and Beckman, 1981; Nelson, 1981). Fusarium oxysporum has no known sexual stage, but produces three types of asexual spores. Macroconidia and microconidia are thin-walled spores that serve to spread the fungus within as well as outside the host plant. Chlamydospores are thick-walled resting structures that can survive for long periods in the soil as dormant propagules in the absence of a suitable host plant (Nelson, 1981).

Fusarium oxysporum f. sp. niveum (E.F.Sm.) Snyd. & Hans., causal agent of Fusarium wilt of watermelon (Citrullus lanatus [Thunb.] Matsum. & Nakai), is pathogenic only on watermelon and occurs throughout the watermelon-growing regions of the world. Because of the severity of this disease and the ability of the pathogen to survive many years in the soil, Fusarium wilt is often a limiting factor in watermelon production.





1








2
Control of Fusarium wilt of watermelon is currently dependent upon the use of effective wilt-resistant cultivars and long rotation periods between watermelon crops (Elmstrom and Hopkins, 1981; Hopkins and Elmstrom, 1984). Resistance to the common race 1 of the pathogen is attributed to a single dominant gene (Henderson et al., 1970; Netzer and Weintall, 1980). Resistance is not complete, however, and forms a continuum ranging from susceptible to highly resistant cultivars. Many of the most highly resistant varieties are not used by commercial growers because of other undesirable characteristics (Elmstrom and Hopkins, 1981). Crop rotation is necessary even with cultivars classified as highly resistant, which often succumb to disease in heavily infested fields (Elmstrom and Hopkins, 1981; Hopkins and Elmstrom, 1984; Hopkins et al., 1987). The recent discovery in the United States of the highly aggressive race 2 (Bruton et al., 1988; Martyn, 1987), to which there is presently no known resistance, represents an additional threat to the commercial cultivation of watermelon. For these reasons current control methods of using available resistant varieties in long rotations are not always effective or practical, and alternative control measures are needed.

Soils that are naturally suppressive to Fusarium wilt diseases of numerous crops are known to occur in many regions of the world (Baker and Cook, 1974; Cook and Baker, 1983; Schneider, 1982; Toussoun, 1975). In these soils, disease does not readily develop even though the pathogen and susceptible hosts are present (Cook and Baker, 1983; Schneider, 1982). Among the most well-known and studied examples of Fusarium wilt-suppressive soils are those of the Chateaurenard region of France, where susceptible vegetables have been grown for centuries with no wilt problems (Alabouvette, 1986; Alabouvette et al., 1979, 1985b; Louvet et al. 1981), and in the Salinas Valley of California, where wilt also does not develop despite years of continuous cropping of susceptible








3
plants (Scher and Baker, 1980, 1982; Smith, 1977; Smith and Snyder, 1972; Sneh et al., 1984; Yuen et al., 1985).

The cause of disease suppression in these and many other such soils has been determined to be biological in origin, with suppressiveness being eliminated with aerated steam treatments at 55-600C for 30 min, fumigation with methyl bromide, or gamma irradiation (Alabouvette, 1986; Cook and Baker, 1983; Louvet et al., 1981; Scher and Baker, 1980). Suppression has often been associated with inhibition of chlamydospore germination and reduced saprophytic growth of the pathogen (Alabouvette, 1986; Alabouvette et al., 1985a, 1985b; Huang et al., 1988; Hwang et al., 1982; Louvet et al., 1981; Scher and Baker, 1980, 1982). These soils generally have many characteristics in common, including their physical-chemical makeup (high pH, high organic matter content, and a high montmorillonite clay content), transmissibility of suppressiveness to certain other soils (as little as 1% suppressive soil needed in some cases), and effectiveness against a number of F. oxysporum formae speciales (Alabouvette, 1986; Alabouvette et al., 1979; Cook and Baker, 1983; Louvet et al., 1979; Schneider, 1982). Biological components from such suppressive soils may provide an alternative or supplemental form of wilt disease control.

Disease suppression in these soils has often been shown to result from specific types of antagonistic organisms acting in conjunction with a large, diverse population of microorganisms supported by the physical characteristics of the soils, thus providing both a specific and general suppressive effect (Alabouvette, 1986; Alabouvette et al., 1985b; Cook and Baker, 1983; Louvet et al, 1981; Scher and Baker, 1980, 1982). The specific suppression in Chateaurenard soils, as well as several other wilt-suppressive soils, has been attributed to nonpathogenic strains of F. oxvsporum and F. solani (Alabouvette,








4
1986; Alabouvette et al., 1985b; Louvet et al., 1981; Schneider, 1984; Tamietti and Pramotton, 1990), whereas suppression in the Salinas Valley and other soils has been more closely associated with certain strains of fluorescent species of Pseudomonas (Elad and Baker, 1985a, 1985b; Scher and Baker, 1980, 1982; Sneh et al., 1984). Numerous other organisms have also been reported as causing or contributing to suppressiveness in various soil systems, including species of Arthrobacter (Smith, 1977; Sneh, 1981), Alcali-qenes (Yuen and Schroth, 1986), Trichoderma (Lin and Cook, 1979; Locke et al., 1985; Marois et al., 1981), Penicillium (Lin and Cook, 1979; Marois et al., 1981), Bacillus (Yuen and Schroth, 1986), Serratia (Sneh, 1981; Sneh et al., 1985), Hafnia (Sneh et al., 1985), and others. Generally, these Fusarium wilt-suppressive soils are considered to be of the long-standing, naturally occurring type (Alabouvette, 1986; Alabouvette et al., 1985b; Cook and Baker, 1983; Hornby, 1983; Louvet et al., 1981; Shipton, 1977); their suppressiveness is inherent in the physical-chemical and microbiological structure, was present before cultivation began, and is independent of any cropping practices of susceptible or resistant plants.

In recent years, Hopkins and co-workers (1985, 1987) have reported the development of a different type of Fusarium wilt-suppressive soil in Florida. This suppressive soil was induced as a result of monoculture of a particular cultivar of watermelon. Out of 10 cultivars with varying levels of resistance to wilt, only cultivar 'Crimson Sweet' promoted the development of suppressive soil with monoculture in the field (Hopkins, 1985; Hopkins et al., 1987). This apparently cultivar-specific induction of a suppressive soil was observed after 2-3 years when 'Crimson Sweet,' normally only moderately resistant to Fusarium wilt in greenhouse and field tests (Elmstrom and Hopkins, 1981; Hopkins and Elmstrom, 1984), showed significantly less wilt and higher








5
yields in the field than any other cultivar. By the 5th or 6th year of monoculture, wilt incidence had reached devastatingly high levels in all other cultivars and had stabilized; disease levels were as great in highly resistant cultivars as in susceptible varieties (Hopkins et al., 1987). The 'Crimson Sweet' field plots, however, have maintained extremely low levels of wilt and high yields in the field for over 10 seasons. Soil from these plots has demonstrated suppressiveness in greenhouse tests using susceptible cultivars and additions of pathogen inoculum (Hopkins et al., 1987).

Previous work with this suppressive soil demonstrated that suppression was biological in origin; it was eliminated by fumigation with methyl bromide as well as with moist heat treatments at 65-700C for 30 min (Hopkins et al., 1987). This suppression occurs in a soil type which is normally highly conducive to disease; it is sandy, has a low pH (6.0-6.5), and a low organic matter content (<1%) (Hopkins et al., 1987). Suppressiveness was not affected by desiccation and was not readily transmissible to other soils; at least 25-50% suppressive soil mixed with conducive soil was required to initiate noticeable suppression (Hopkins and Larkin, unpublished). Other preliminary experiments indicated adjustments in the pH of the soil from pH 4 to pH 8 had no effect on suppression and selective elimination of portions of the microflora by bacterial antibiotic treatments (including penicillin, polymyxin, streptomycin, and vancomycin) also did not significantly affect suppression (Hopkins and Larkin, unpublished). The organisms responsible for this suppression, as well as the mechanisms involved, have not yet been identified.

The mechanisms active in suppressive soils are still not completely understood. Components in the soil may suppress the pathogen directly by the destruction of hyphae or propagules by lysis or parasitism, by inhibiting propagule germination, or by reducing








6
saprophytic growth in some other way (Baker and Cook, 1974; Cook and Baker, 1983). Most suppressive soils have a higher level of fungistasis than conducive soils, with chlamydospore germination and germ tube length of the pathogen greatly reduced (Alabouvette, 1986; Alabouvette et al., 1979, 1985b; Huang et al., 1988; Hwang et al., 1982; Louvet et al., 1981; Smith, 1977; Smith and Snyder, 1972; Sneh et al., 1984). Antagonistic microorganisms may interfere with the nutrient supply to the pathogen and effectively prevent infection. Nonpathogenic strains of F. oxysporum have been suggested to function in suppression by effectively competing with the pathogen, either saprophytically for nutrients in the rhizosphere (Alabouvette, 1986; Alabouvette et al., 1985b; Louvet et al., 1981) or parasitically for infection sites on the root (Schneider, 1984). Fluorescent pseudomonads produce a siderophore that complexes iron in the rhizosphere, making it unavailable to the pathogen, and is thought to be the mechanism of suppression in some soils (Elad and Baker, 1985a, 1985b; Kloepper et al., 1980; Kloepper and Schroth, 1981; Scher and Baker, 1980, 1982). Some fluorescent pseudomonads also produce phenazine antibiotics, which may also be important factors in suppression in some systems (Hamdan et al., 1988; Pierson and Thomashow, 1988; Thomashow et al., 1988; Thomashow and Weller, 1988). However, an active total microbial population providing intense competition for nutrients also seems to be important in many soils. Suppressiveness also may act indirectly through the induction of host resistance to the pathogen caused by prior colonization by certain soil microorganisms. Induced resistance to F. oxvsporum f. sp. batatas has been demonstrated in sweet potatoes previously inoculated with nonpathogenic strains of F. oxysporum (Ogawa and Komada, 1984,1985,1986). Numerous other studies have shown that nonpathogenic or avirulent strains of F. oxysporum applied to roots can protect the








7
host from disease when challenged by a virulent strain (Biles and Martyn, 1989; Davis, 1967, 1968; Gessler and Kuc, 1982; Martyn et al., 1990; Wymore and Baker, 1982). It is also possible that more than one mechanism may be operating in a particular suppressive soil.

Various biological components from suppressive soils have been and are being studied for their potential to control Fusarium wilt and other diseases. Specific antagonistic organisms have been isolated, screened, and tested for their ability to reduce disease in the greenhouse and in the field. However, because of the complex nature of such suppressive soils, in addition to identifying the organisms responsible, it is important to study the ecology of these soils and the specific conditions, processes, interactions, and mechanisms which make them suppressive (Baker and Chet, 1982; Cook and Baker, 1983; Toussoun, 1975). A thorough understanding of the ecological interactions of the pathogen and other microorganisms may be vital to the effective utilization of specific antagonists from these soils as biological control agents.

The unique features of the 'Crimson Sweet' suppressive monoculture soil from Florida provide an excellent opportunity for analyzing the ecological characteristics, organisms, and mechanisms responsible for suppression in this soil system. The cultivarspecific induction of suppressiveness and its development in a soil type which is normally highly conducive to wilt distinguishes this soil from most other Fusariumn wilt-suppressive soils, and makes this soil especially applicable for comparative analysis with similar conducive soils.

The overall objectives of this research were to determine the nature of the microbiological basis of this suppressive soil and to provide a better understanding of how this cultivar-induced suppressive soil works. This work was undertaken as the initial








8

step in the development of a potential biological control system incorporating components or mechanisms from this soil. The more specific objectives were to (1) evaluate various ecological characteristics of the suppressive soil in relation to the pathogen, (2) identify the organism(s) responsible for this suppression, and (3) identify the mechanism(s) responsible.

In the past, ecological and epidemiological studies with F. oxvsr~orum have been severely limited by the fact that particular pathogenic formae speciales are morphologically indistinguishable from other formae speciales and saprophytic strains of F. oLxvsporum which occur abundantly in most soils. In this study, precise identification, differentiation, and clarification of the different races of the pathogen, as well as gaining an understanding of their distribution and role in suppressive and conducive soil, may be important for understanding the mechanism of suppression. Therefore, an additional preliminary objective of this research was to develop a relatively quick, accurate, and efficient technique for identifying and distinguishing strains of F. oxvsporum f. sp. niveum from those not pathogenic to watermelon, as well as differentiating pathogenic races within F. oxvsrorum f. sp. niveum.













CHAPTER 2
VEGETATIVE COMPATIBILITY WITHIN FUSARIUM OXYSPORUM
F. SP. NIVEUM AND ITS RELATIONSHIP TO VIRULENCE, AGGRESSIVENESS, AND RACE


Introduction


Fusarium oxvsporum Schlecht. f. sp. niveum (E.F.Sm.) Snyd. & Hans., the causal organism of Fusarium wilt of watermelon (Citrullus lanatus [Thunb.] Matsum. & Nakai), is widespread throughout the watermelon-growing regions of the world. This forma specialis is pathogenic only on watermelon and has been subdivided into two or three pathogenic races according to virulence on cultivars of varying levels of resistance (Crall, 1963; Elmstrom and Hopkins, 1981; Netzer, 1976; Netzer and Weintall, 1980). However, these races are not clearly defined and there has been difficulty in their differentiation (Armstrong and Armstrong, 1978; Martyn, 1987; McKeen, 1951).

Crall (1963) originally described two races of F. oxysporum f. sp. niveum in Florida; following the numbering system of Cirulli (1972), these are designated races 0 and 1. In general, both races caused severe wilt on susceptible cultivars, but race 0 did not wilt resistant cultivars while race 1 caused slight to moderate wilt on most cultivars classified as resistant to Fusarium wilt. Armstrong and Armstrong (1978), however, concluded that differences between these strains were not great enough to constitute distinct races and considered them all to be race 1. In 1976, Netzer (1976) described a highly aggressive race in Israel capable of causing severe wilt in all known resistant



9








10
cultivars; it was designated race 2. Within the last few years race 2 has been reported in the United States, but only in Texas (Martyn, 1987) and Oklahoma (Bruton et al., 1988).

Traditionally, F. oxvsporum f. sp. niveum can be distinguished from other formae speciales and saprophytic strains of F. oxysporum only by its virulence on watermelon. Distinguishing races require screening of pathogenic isolates on cultivars of varying levels of resistance. These tests are laborious and often inconsistent or inconclusive. Results can be greatly influenced by environmental factors, host age, inoculum level, and inoculation methods (Bosland et al., 1988; Latin and Snell, 1986; Martyn and McLaughlin, 1983).

An alternative approach to the classification of strains of F. oxvsporum is based on vegetative compatibility (Correll et al., 1987; Puhalla, 1985). Strains that are capable of forming heterokaryons with each other are vegetatively compatible and all such compatible strains comprise a distinct vegetative compatibility group (VCG). Since F. oxysporum lacks a sexual stage, genetic interaction is thought to be restricted primarily to strains that are vegetatively compatible with one another. Puhalla (1985) suggested that each VCG may represent a genetically isolated population, resulting in specific genetic traits being associated with particular VCGs. Using complementary nitrate-nonutilizing mutants to detect hetrokaryon formation, Puhalla demonstrated that particular VCGs always correspond with formae speciales and often with races within formae speciales. Subsequent studies have used vegetative compatibility to define, subdivide, or differentiate strains and pathogenic races in F. oxysporum f. sp. ?LH (Correll et al., 1986a; Ireland and Lacy, 1986), f. sp. pjaj (Correll et al., 1985), f. sp. conglutinans (Bosland and Williams, 1987), f. sp. cubense (Ploetz and Correll, 1988), f. sp. melonis (Jacobsen and Gordon, 1988), f. sp. lycopersici (Elias and Schneider, 1987), f. sp. vasinfectum (Katan










and Katan, 1988), f. sp. dianthi (Katan et al., 1989), and f. sp. asgaragi (Elmer and Stephens, 1989). It has also been used to differentiate nonpathogenic forms (Correll et al., 1 986b; Elias and Schneider, 1987).

The objectives of this study were to determine the potential for using vegetative compatibility to distinguish F. oxvsoorum f. sp. niveum from other strains of F. oLxvsporum as well as in the differentiation or clarification of pathogenic races or other subdivisions within F. oLxvs~orum f. sp. niveum.

Because terms such as "virulence" and "aggressiveness" have been used differently by various authors, it is necessary to clarify how these terms have been defined here. In this paper, the concepts of virulence and aggressiveness as discussed by Vanderplank (1975, 1978) and Caten (1987) are used. Virulence is considered a qualitative trait of the pathogen associated with vertical resistance in the host. Thus, virulence refers to the ability of a pathogen to cause disease on a particular host cultivar and is related to race. Aggressiveness, on the other hand, refers to the level or degree of pathogenicity, a quantitative trait, measured by disease severity on a number of different host cultivars. Aggressiveness varies independently of the differences in host cultivars and is related to horizontal resistance. Also, for the purposes of this paper, all references to pathogens or pathogenicity relate only to the ability to cause Fusarium wilt in watermelon and not to pathogens causing disease on any other host. A preliminary report of this work has been published (Larkin et al., 1988).








12
Materials and Methods


Isolates of Fusarium oxysorum


Initial tests for establishing vegetative compatibility groups were made using Florida isolates from a previous study (Hopkins et al., 1987), which had been tested for pathogenicity and identified as either F. oxysporum f. sp. niveum or miscellaneous isolates of F. oxvsporum not pathogenic on watermelon. These included 27 isolates of F. oxvsporum f. sp. niveum collected from stems of naturally infected watermelon plants as well as 25 isolates of F. oxysporum collected from field plot soils in Florida. For the present study, an additional 180 isolates of F. oxvsporum were collected from the same field plot soils. Isolates of F. oxysporum f. sp. niveum originating from various locations around the world were obtained from individuals and culture collections (Table 2-1).

Fungal isolations were made from soil by adding a 5.0-g subsample to 45 ml sterile water and stirring vigorously for 10 minutes. While being stirred, 1 ml of this suspension was extracted and added to 9 ml 0.1 % water agar and agitated using a vortex mixer. One milliliter of this suspension was pipetted onto each of five plates of Komada's (1975) Fusarium-selective medium and spread over the surface by tilting the plates. Isolates were identified as F. oxvsporum by their characteristic growth and morphology on this medium 7-10 days later. All isolates were single-spored and cultured at 260C on potato dextrose agar (PDA) or a minimal medium (MM) containing nitrate as its only nitrogen source (Puhalla, 1985). Isolates were stored on sterile filter paper at 4C (Correll et al., 1986a_).








13
Recovery of Nitrate-Nonutilizing Mutants


Methods similar to those developed by Puhalla (1985) and Correll et al. (1987) were used. A mycelial transfer (8-mm3 PDA block) of each isolate of F. oxvsporum was placed on a chlorate-amended medium of either PDA or MM containing 1.5% KCIO3 (Puhalla, 1985). Initially, colonies were greatly restricted, but after a few days fast-growing sectors emerged from most colonies. Portions of these sectors were transferred to MM. Colonies producing a characteristically thin expansive growth with no aerial mycelium on this medium were considered to be nitrate-nonutilizing (nit mutants. All nit mutants were resistant to chlorate and showed wild-type growth on PDA. Numerous mutants were generated for each isolate. When paired on MM, complementary mutants were identified by the production of a distinct line of thick wild-type growth where the two colonies came in contact. This was the visual indication of hyphal anastomosis and heterokaryosis.


Vegetative Compatibility Tests


To establish VCGs, two complementary nit mutants derived from each isolate were paired in all combinations with complementary mutants of other isolates. Pairings were made by placing mycelial blocks 1.5 cm apart on MM and incubating for 7-14 days at 261C. Compatible isolates produced a heterokaryotic tuft of wild-type growth between complementary mutants. Incompatible isolates produced no reaction. After distinct vegetative compatibility groups were established, two pairs of complementary nit mutants were selected from each VCG for use as tester strains and were used to screen all subsequent isolates for compatibility. Mutants to be used as testers were classified according to their nitrogen utilization phenotype. Each tester pair consisted of one NitM and one nit1 or nit3 phenotypic mutant as designated and defined by Correll et al.(1 987).








14

To screen numerous isolates quickly and efficiently, a multiple-cross plating technique was used to maximize the number of crosses per plate. Using a color-coded template, six different isolates could be crossed with up to seven tester isolates in all combinations. All pairings were made at least twice using two different tester strains from each VCG.


Pathogqenicity Tests and Race Determinations


Pathogenicity tests were conducted with all isolates. Initial tests were conducted using susceptible watermelon cultivars ('Florida Giant' or 'Sugar Baby') to separate pathogenic from nonpathogenic isolates. Subsequent virulence tests with pathogenic isolates to determine race designations used six cultivars, ranging in order from susceptible to highly resistant as follows: 'Florida Giant,' 'Charleston Gray,' 'Crimson Sweet,' 'Sugarlee,' 'Dixielee,' and 'Calhoun Gray.' Race determinations were made on all imported isolates and selected Florida isolates representative of each VCG. Inoculum consisted of microconidial suspensions harvested from 5- to 7-day-old PDA cultures. Conidia were added at the rate of 5 x 1 03/g field soil which had been microwave-irradiated (2450 MHz, 700 watts) for 2 min/kg soil at a matric potential of -0.01 MPa. This treatment was effective in eliminating indigenous F. oxysporum propagules in preliminary tests. The soil used in these tests was of the Chipley sand soil series, which are thermic, coated Aquic Quartzipsamments; it had a pH of about 5.5, cation exchange capacity of 12-16 meq/1 00 g soil, and an organic matter content of about 2% in the surface layers. Seeds were planted directly into the infested soil. Initial pathogenicity tests consisted of four to six replicate pots (7.5-cm diameter) of six plants each. Race determination tests used four replicate pots (10-cm diameter) of 10 plants each. Plants were maintained in a








15
greenhouse at 21-320C, where maximum light intensity was 500-700 Pmol/m2Is. Wilted seedlings were counted and removed weekly for 4 weeks and the percent wilt incidence was calculated. All tests were conducted at least twice.


Results


Vegetative Compatibility Tests


All isolates of F. oxysporumn f. sp. niveum were found to belong to one of three distinct VCGs. Following the VOG numbering system adopted by Puhalla (1985) these are designated 0080, 0081, and 0082 (Table 2-1). Isolates which were not pathogenic on watermelon were incompatible with all isolates pathogenic on watermelon. All isolates within a VCG produced unambiguous compatible reactions with all other isolates of that VCG and no isolates were compatible with isolates from more than one VOG. All pathogenic isolates were self-compatible (Jacobsen and Gordon, 1988).

All of the imported isolates, with the exception of the two isolates of race 2 from Texas, were found to be within VOG 0080 regardless of geographic origin (Table 2-1). This VOG also contained many Florida isolates collected from both wilted plants and soil samples. Testing of isolates obtained from J.M.Crall showed an isolate originally designated by him as race 0 (JMVC6O-3) to be in VCG 0080, while isolates Crall classified as race 1 were contained in VCGs 0081 and 0082. No imported isolates were vegetatively compatible with VOG 0081, which to date consists only of isolates from Florida. Race 2 isolates from Texas were found to be in the same compatibility group (VCG 0082) as many highly aggressive Florida strains.

In screening 180 isolates of F. oxvsporum from soil for vegetative compatibility, only 12 isolates were found to be compatible with one of the three established pathogenic VOGs. Ten of these were in VCG 0081 and 1 each in VCGs 0080 and 0082.








16
Pathocienicily Tests


All but two isolates that had been identified previously as F. oxvYsporum f. sp. niveum were verified to be pathogenic on watermelon, producing 50-100% wilt in susceptible cultivars. The two nonpathogenic isolates were not compatible with any of the three pathogen VCGs. Nonpathogenic isolates were incapable of systemically infecting and causing wilt in watermelon. Of the 180 isolates of F. oxvsporum from soil, only the 12 isolates that were compatible with the pathogen VCGs were pathogenic on watermelon, causing 15-85% wilt in susceptible cultivars. Thus, all pathogenic isolates and only pathogenic isolates were contained in VCGs 0080, 0081, and 0082.


Race Determination


All pathogenic isolates were capable of causing substantial wilt (49-100%) in the susceptible cultivar 'Florida Giant,' but varied greatly in the amount of wilt produced in resistant cultivars; (Table 2-2). Combined results of two separate virulence tests (individual test results are presented in Appendix A; Tables A-i and A-2) with the imported isolates demonstrated the wide range of aggressiveness (7-90% wilt) observed on cultivars classified as resistant. Isolates produced varying levels of wilt on 'Charleston Gray' and 'Crimson Sweet,' with considerable overlap among isolates and no distinct differences attributable to race. Distinctions between races could only be observed on the most highly resistant cultivars. Texas race 2 isolates TX-X1 D and TX-HC3 and Florida isolate CS85-4 were all vegetatively compatible and produced comparably severe levels of wilt on all cultivars, with significantly greater wilt (67-75%) on 'Dixielee' and 'Calhoun Gray' than with all other isolates tested. Isolates FG85-2 and FG85-1 5, both in VCG 0081, produced comparable wilt to some of the race 2 isolates, as well as significantly greater








17

wilt than the remaining race 1 isolates, on 'Sugarlee,' but not on 'Dixielee' or 'Calhoun Gray.'

An additional virulence test was conducted using several Florida isolates representing the three VCGs. Although generally higher levels of wilt were observed across all VCGs than in previous tests, similar differences regarding race were confirmed (Appendix A; Table A-3). The two isolates in VCG 0082 produced significantly greater wilt than all other isolates on 'Calhoun Gray.' Additional pathogenicity tests of other VCG 0082 isolates on 'Calhoun Gray' confirmed their ability to induce severe wilt on highly resistant cultivars (60-100%). Based on these virulence tests and vegetative compatibility data, the numerous Florida isolates in VCG 0082 were comparable in all respects to the race 2 isolates from Texas and were thus considered to be race 2.

Race 2 isolates separated distinctly from race 1 isolates on 'Calhoun Gray' in all tests. No other distinctions of race could be clearly differentiated, so all other isolates were considered to be race 1. In a separate test involving an isolate previously designated as race 0 (JMC6O-3), no clear distinction could be made between the race 0 isolate and other weakly aggressive race 1 isolates (data not shown).


Relationship of Vegetative Compatibilitv Groups to Aggressiveness and Race


Isolates in VCG 0080 showed a wide range of aggressiveness, producing from fairly low to substantial wilt in resistant cultivars, yet all were distinguishable from race 2 isolates (Table 2-2). Isolates in VCG 0081 generally produced higher levels of wilt than VCG 0080 isolates, but there was some overlap with many VCG 0080 isolates. All VCG 0082 isolates caused severe wilt in all cultivars; and were considered to be race 2.








18

Comparing the virulence data by VCG (Table 2-3) clarified this association of VCG to aggressiveness and race. Isolates in VCG 0082 averaged significantly greater wilt than did VCG 0080 and VCG 0081 isolates on 'Crimson Sweet,' 'Dixielee,' and 'Calhoun Gray' in all tests, and on 'Charleston Gray' in the imported isolate tests. Vegetative compatibility group 0081 wilt averages were significantly different from those of both VCG 0080 and VCG 0082 on 'Calhoun Gray' and 'Dixielee,' demonstrating a level of aggressiveness intermediate to that of race 2 and VCG 0080 race 1 isolates. However, in the Florida isolate test (Table 2-3), VCG 0081 wilt averages also were significantly lower (less aggressive) on 'Crimson Sweet' and 'Charleston Gray' than those of VCG 0080 and 0082. This difference in aggressiveness between VCG 0080 and VCG 0081 varied from test to test and with the isolates used, and did not appear to be consistently distinct or sufficiently reliable to constitute a race difference. Thus, as defined in this study, race 1 isolates were contained in two distinct VCGs, 0080 and 0081, while all race 2 isolates were in VCG 0082 (Table 2-4).


Discussion


Results from this study suggest a direct relationship between vegetative compatibility groups and virulence in F. oxysrorum f. sp. niveum. All isolates of F. oxysporum f. sp. niveum studied were contained within three distinct VCGs, whereas isolates not pathogenic on watermelon were excluded from these VCGs. Vegetative compatibility tests were used to effectively identify isolates of F. oxysporum f. sp. niveum from within a population of miscellaneous strains of F. oxvsporum; these findings were verified by pathogenicity tests. Although previously described races in F. oxsporum f. sp. niveum, such as race 0, could not always be differentiated by virulence tests, vegetative








19


Table 2-1. Isolates of Fusarium oxysporum f. sp. niveum used in comparative pathogenicity tests.

Isolate Origin Race VCG Source

JMC60-3 Florida (0)1- 0080 J.M.Crallb
0-936 Maryland 1 0080 P.Nelsonc
0-987 Maryland 1 0080 P.Nelson
O-1210 Indiana 1 0080 P.Nelson
O-1128 California 1 0080 P.Nelson
0-1130 California 1 0080 P.Nelson
O-1132 Taiwan 1 0080 P.Nelson
0-1182 Australia 1 0080 P.Nelson
0-974 Australia 1 0080 P.Nelson
18467 South Carolina 1 0080 ATCCd
44293 California 1 0080 ATCC
FG85-1 Florida 1 0080 Locale
CS85-1 Florida 1 0080 Local
FG85-2 Florida 1 0081 Local
FG85-15 Florida 1 0081 Local
FG85-20 Florida 1 0081 Local
JMC72-19 Florida 1 0081 J.M.Crall

CS85-4 Florida 2 0082 Local
CG85-15 Florida 2 0082 Local
TX-X1 D Texas 2 0082 R.D.Martyn'
TX-HC3-13B Texas 2 0082 R.D.Martyn
JMC70-1A Florida 2 0082 J.M.Crall
JMC72-8 Florida 2 0082 J.M.Crall

"This isolate was originally designated race 0 by J.M.Crall, but is considered to be race 1 in the present study.
b Central Florida Research and Education Center, Leesburg, FL
Fusarium Research Center, Pennsylvania State University, University Park, PA
d American Type Culture Collection, Rockville, MD
e Author, Leesburg, FL
Texas A & M University, College Station, TX








20

Table 2-2. Wilt development in different watermelon cultivars planted in soil infested with various isolates of Fusarium oxysporum f. sp. niveum.

Percent wilt on indicated cultivars
Isolate VCG Race FGb ChG CS SL DL CalG Mean

18467 0080 1 60ac 21 a 37a 9a 8a 7a 24
O-1132 0080 1 62ab 26a 42abc 10Oab 16ab 11a 28
0-1128 0080 1 74bc 32ab 45abcd 17abc 19ab 16ab 33 0-936 0080 1 91 de 58def 43abc 23abcd 20ab 13a 41 O-1182 0080 1 78c 46bcd 41abc 29cde 37cd 18abc 41 0-987 0080 1 83cd 58def 51abcd 26abcde 25bc 18abc 43 FG85-1 0080 1 95de 53cde 50abcd 41e 17ab 14a 45
O-1130 0080 1 84cd 54cde 53bcd 33cde 30bcd 17abc 45 O-1210 0080 1 94de 62def 57cde 28bcde 30bcd 30cd 51 0-974 0080 1 89e 68ef 60de 38de 50d 27bc 56

FG85-2 0081 1 95de 38abc 42abc 57f 47d 49e 55
FG85-15 0081 1 95de 51cde 39ab 70fg 45d 42de 57

TX-X1 D 0082 2 85cd 68ef 58de 76g 69e 74f 72
TX-HC3 0082 2 100e 78fg 74ef 71fg 70e 68f 77 CS85-4 0082 2 94de 90g 87f 83g 67e 75f 82

Cultivar Averages 85 52 51 40 36 31
a Percent wilt indicates the incidence of Fusarium wilt after 4 weeks.
b Cultivars used were 'Florida Giant'(FG), susceptible to Fusarium wilt; 'Charleston Gray'(ChG) and 'Crimson Sweet'(CS), moderately resistant; 'Sugarlee'(SL), 'Dixielee'(DL), and 'Calhoun Gray'(CalG), highly resistant.
C Means within columns followed by the same letter are not significantly different (<0.05) according to Duncan's multiple range test. Means in each of two tests were based on four replicate pots of 10 plants each for each isolate/cultivar combination. Results of the two similar tests were combined for analysis (Individual test results are in Appendix A; Tables A-1 and A-2). Analysis was conducted on actual incidence data. Inoculum consisted of conidia added to the soil at 5 x 103/g soil.








21

Table 2-3. Average percent wilt in various watermelon cultivars caused by isolates of Fusarium oxvsporum f. sp. niveum in three vegetative compatibility groups.

Percent wilt on indicated cultivars
VCG Race FGb ChG CS SL DL CalG Mean

Imported isolate tests

VCG 0080 1 82ad 46a 48a 28a 25a 17a 41
VCG 0081 1 95a 44a 40a 63b 46b 45b 56
VCG 0082 2 91a 79b 75b 83b 70c 73c 78

Florida isolate tested

VCG 0080 1 96a 86b 76b 75ab 54a 37a 71
VCG 0081 1 93a 59a 64a 72a 64b 49b 67
VCG 0082 2 99a 96b 90c 87b 82c 86c 90
a Percent wilt indicates the incidence of Fusarium wilt after 4 weeks.
b Cultivars used were 'Florida Giant'(FG), susceptible to Fusarium wilt; 'Charleston Gray'(ChG) and 'Crimson Sweet'(CS), moderately resistant; 'Sugarlee'(SL), 'Dixielee'(DL), and 'Calhoun Gray'(CalG), highly resistant.
c Values represent the combined averages of two similar tests (Individual test results are in Appendix A; Table A-4). Isolates used were ATCC 18467, 0-936, 0-974, O978, O-1128, O-1130, O-1132, O-1182, O-1210, and FG85-1 (all VCG 0080); FG85-2 and FG85-15 (VCG 0081); and TX-X1 D, TX-HC3, and CS85-4 (VCG 0082).
d Means within columns in each test followed by the same letter are not significantly different (p<0.05) according to Duncan's multiple range test. Means were based on four replicate pots of 10 plants each for each isolate/cultivar combination. Analysis was conducted on actual incidence data. Inoculum consisted of conidia added to the soil at 5 x 1 03/g soil.
Florida isolate test consisted of isolates FG85-1 and CS85-1 (VCG 0080); FG852, FG85-15, FG85-20, and JMC72-19 (VCG 0081); CS85-4 and CG85-15 (VCG 0082). Same inoculum level and procedures were used as in the previous test.








22





Table 2-4. Vegetative compatibility groups and characteristics established for Fusarium oxvsporum f. sp. niveum. VCG No.isolates Race Origin 0080 19 1 California, Florida, Indiana,
Maryland, South Carolina, Australia, Taiwan.

0081 17 1 Florida

0082 24 2 Florida, Texas

Total 60








23

compatibility groups did correspond with the two races that were identified. Within each VCG there existed a range of aggressiveness, yet there were distinct differences in wilt averages when grouped by VCG, with a progression of increasing aggressiveness from VCG 0080 to 0082.

Difficulties in the identification and differentiation of strains of F. oxysporum have long been major limitations to ecological and epidemiological studies with these important soil organisms. The genus Fusarium is well known for its variation within species and subspecies with respect to pathogenicity as well as morphology (Armstrong and Armstrong, 1978; Toussoun and Nelson, 1975). This variability, combined with the added variability and inconsistency associated with virulence tests, has rendered tenuous the classification of pathogenic races within many formae speciales.

For many pathogenic fungi, races are defined on the basis of differential cultivars, which have different genes for resistance to specific pathogenic races. However, races within F. oxvsporum f. sp. niveum have not been this clearly defined. Early investigations (McKeen, 1951; Reid, 1958; Sleeth, 1934) reported variability in aggressiveness among isolates without any clear indication of distinct races. Even when races were identified, race differentiation depended upon gradations in the degree of wilt produced on various cultivars rather than unambiguous susceptible or resistant reactions (Cirulli, 1972; Crall, 1963; Martyn, 1987). Because of this, it has been suggested that true races do not exist in this forma specialis, but merely populations with variable aggressiveness (Armstrong and Armstrong, 1978; McKeen, 1951). Results of this study tend to support this contention. Differences between races were primarily those of level of aggressiveness on particular cultivars, rather than in virulence on differential cultivars. All pathogenic isolates appeared to be virulent on all cultivars, with differences only in the degree of wilt








24
produced. Caten (1987) considered racial classification by level of aggressiveness to be ambiguous and impractical. However, there do appear to be sufficient differences in aggressiveness between races 1 and 2 to justify continued classification into distinct groups whether or not these are classified as true races. The association of VCGs with these distinct pathogenic groups suggests that there are genetic differences between the groups and that VOGs can be used to identify and differentiate them. Two distinct groups within race 1 which could not be clearly distinguished by virulence tests were differentiated by vegetative compatibility.

Results from the virulence tests in this study demonstrated some of the problems associated with race identification. Isolates classified as race 1 caused from 0-50% wilt on cultivars considered resistant to race 1. Isolate ATCC 18467 was used in this study and also by others (Armstrong and Armstrong, 1978; Marlyn, 1987) as a representative of race 1. In our tests ATCC 18467 was one of the most weakly aggressive of the race 1 isolates tested, and thus may not be the best isolate to use for comparison. It may be necessary to include several other race 1 isolates in such comparisons to account for the variation of wilt reactions found within this race (Martyn, 1987).

Martyn (1987) reported that several cultivars, including 'Calhoun Gray' and 'Crimson Sweet,' could be used to differentiate races 1 and 2. In the present study, 'Dixielee' and 'Calhoun Gray' were the best cultivars for separating races 1 and 2, whereas 'Crimson Sweet' was not very effective. Both 'Charleston Gray' and 'Crimson Sweet' appeared to be susceptible to some race 1 isolates but resistant to others. Marlyn (1987) also noted that no cultivar was wholly effective in differentiating race 0 from race 1. Likewise, in this study no distinction could be made between races 0 and 1. The only isolate tested which was purported to be race 0 could not be distinguished from several








25

race 1 isolates. Cirulli (1972) based his distinction between these two races on a differential response to 'Charleston Gray;' such a response was not evident in this study. Although two subgroups were defined within race 1 (VCG 0080 and VCG 0081) and showed some differences in wilt reactions on resistant cultivars, these differences were not always clear and did not correspond with past descriptions of race 0. There were no isolates which were unable to cause some degree of wilt on resistant cultivars (race 0) (Cirulli, 1972; Crall, 1963). The distinction between race 0 and 1 has been disputed by others (Armstrong and Armstrong, 1978; Martyn, 1987) and there is no evidence here for classification as separate races.

Some of the differences observed between VCGs 0080 and 0081 also may be misleading. Although in some tests isolates in VCG 0081 were more aggressive than other race 1 isolates (Tables 2-2 and 2-3), these differences were based on too few isolates and do not indicate the entire range of aggressiveness observed in VCG 0081. Some of the isolates in VCG 0081 collected from the soil (not used in race determination tests) were only weakly aggressive. Overall, this VCG showed a wide range of aggressiveness among isolates. More extensive testing is needed to determine any real differences in aggressiveness between these two VCGs.

The inoculation method used in this study differed from that used in other studies (Latin and Snell, 1986; Martyn, 1987; Martyn and McLaughlin, 1983), but inconsistensies resulting from this procedural difference appeared to be minimal. In this study, soil was infested with conidial suspensions and then watermelon was seeded directly into the infested soil. This method was faster and easier than the root-dip transplant method commonly used. Since watermelons are usually direct-seeded in the field, no additional wounding or transplant shock was given to the test plants. Also, this inoculation method








26
allowed the pathogen to attack from the very earliest stages of development, as it would in the field. Additional tests were conducted using chlamydospores as inoculum and results comparable to those reported here were obtained (unpublished).

Virulence test results and cultivar resistance rankings from this study are in general agreement with those previously reported (Elmstrom and Hopkins, 1981; Hopkins and Elmstrom, 1984; Martyn, 1987; Martyn and McLaughlin, 1983). However, these results differ from those of Martyn (1987) regarding the resistance of 'Crimson Sweet.' In Martyn's tests, 'Crimson Sweet' consistently rates as one of the most highly resistant cultivars, equalling or surpassing 'Calhoun Gray.' In tests in this study as well as in numerous previous tests (Elmstrom and Hopkins, 1981; Hopkins and Elmstrom, 1984; Hopkins et al., 1987), 'Crimson Sweet' rates as moderately resistant (comparable with the resistance of 'Charleston Gray') in greenhouse and laboratory tests using fumigated or microwavetreated field soil. However, a much higher resistance in field situations and in field soil has been observed (Hopkins and Elmstrom, 1984; Hopkins et al., 1987). Since this appears to be a consistent and repeatable difference which only occurs in this one cultivar, it may be related to the unique ability of 'Crimson Sweet' to promote soil suppressiveness (Hopkins et al., 1987). Perhaps, Martyn's (1987) testing procedure in some way incorporates this additional level of resistance through the development of organisms active in the suppressiveness promoted by 'Crimson Sweet.' This may be related to the background microbial populations present or the type of soil medium and seedling transplant method used.

Numerous race 2 isolates were found to occur in Florida, and this is the first report of this race in the state. Previously, race 2 had been identified in the U.S. only in Texas and Oklahoma (Bruton et al., 1988; Martyn, 1987). Most of the race 2 strains in Florida








27

were isolated from field plots in which resistant cultivars had been grown in a monoculture for several years (Hopkins et al., 1987). A high percentage (>90%) of the pathogens isolated from wilted plants in these plots were race 2 (VCG 0082) isolates. It appears that growing resistant cultivars will select for and increase the abundance of race 2 propagules relative to race 1 over time (Hopkins et al., 1989; Hopkins and Lobinske, 1990). Two of the isolates found to be in VCG 0082 were isolates of J.M.Crall that were collected in 1970 and 1972. Based on this and other reports of higher than normal wilt in resistant cultivars in the field, race 2 has probably been present in Florida in low populations for many years. This may explain why even resistant cultivars must be grown in long rotations in Florida (Elmstrom and Hopkins, 1981; Hopkins and Elmstrom, 1984; Hopkins et al., 1987).

Compared to other formae speciales in which vegetative compatibility has been studied, F. oxysporum f. sp. niveum appears to be one of the least complex, having few VCGs and a direct relationship between VCG and race. Thus, characterization by vegetative compatibiltiy can be implemented and determined with relative ease. Other F. oxysporum formae speciales, including f. sp. gpii (Correll et al., 1986a), f. sp. conglutinans (Bosland and Williams, 1987), f. sp. vasinfectum (Katan and Katan, 1988), and f. sp. dianthi (Katan et al., 1989), also have small numbers of VCGs and relatively direct associations with race. There are some formae speciales which have numerous VCGs and more complex relationships with race, such as F. oxvsporum f. sp. asparapi (Elmer and Stephens, 1989), f. sp. cubense (Ploetz and Correll, 1988), f. sp. lycopersici (Elias and Schneider, 1987), and f. sp. melonis (Jacobsen and Gordon, 1988). It appears the utility of VCGs for identification and differentiation depends on the particular formae speciales, because each has a different relationship determined by the number of VCGs,








28

races, and genetic lines involved. This may be a function of the age of the pathosystem. Ancient pathosystems have had a longer time to evolve, and may result in increased genetic diversity. Similarly, VOGs which are more widely distributed geographically are considered to be older than those which only occur in limited regions. If this is the case, it would appear that VCG 0080 may be the oldest and that VCGs 0081 and 0082 may have arisen more recently.

Within F.,oxvsporum f. sp. niveum, vegetative compatibility can be utilized as an alternative, or at least collaborative, method to distinguish pathogenic from nonpathogenic strains of F. oxvsporum and to differentiate sub-forma specialis virulence characteristics. Vegetative compatibility groups were closely associated with aggressiveness and race characteristics. This supports Puhalla's (1985) hypothesis that VCGs define distinct groups with particular genetic characteristics and have important biological and pathological significance. Because of the problems in virulence tests and the tenuous nature of race relationships, vegetative compatibility groupings may allow more precise, efficient, and objective distinctions, and may provide more appropriate divisions for subdividing F. oxvs~orum f. sp. niveum than those based solely on virulence reactions.














CHAPTER 3
ECOLOGY OF FUSARIUM OXYSPORUM F. SP. NIVEUM IN SOILS
SUPPRESSIVE AND CONDUCIVE TO FUSARIUM WILT OF WATERMELON


Introduction


Because of the complex nature of soils suppressive to plant disease, it is important, in addition to identifying the organisms responsible, to study the overall conditions, processes, and mechanisms which make them suppressive (Baker and Chet, 1982; Cook and Baker, 1983; Louvet, 1989; Toussoun, 1975). A thorough understanding of the ecological interactions of the pathogen and other microorganisms may be vital for the effective utilization of specific antagonists from these soils as biological control agents. Through the analysis of environmental differences in suppressive and conducive soils, much insight may be gained into the effect of the suppressive soil on the pathogen and on how suppression occurs. This information could then be used to study specific interactions, organisms, and mechanisms important in the suppressive response. Only after such ecological characteristics and the active mechanisms are known can strategies be developed to enhance the antagonistic interactions under the proper conditions and at the appropriate sites to enable biological control to function to its fullest.

The soil used in this study is suppressive to Fusariumn wilt of watermelon and was developed through monoculture of watermelon cultivar 'Crimson Sweet.' The cultivarspecific induction of suppressiveness in this soil and its development in a soil type which is normally highly conducive to wilt, makes it especially applicable for comparative


29








30

analysis with similar conducive soils. Using four soils to represent different suppressive and conducive conditions, the objectives of this study were to monitor the population dynamics and chlamydospore germination of Fusarium oxvsporum f. sp. niveum and to evaluate the colonization of watermelon roots by F. oxysporum in relation to other microorganism populations and the incidence of Fusarium wilt. A preliminary report of portions of this work has been published (Larkin et al., 1989).


Materials and Methods


Soils Used


Four soils were used throughout this research to represent different conditions of suppressiveness and conduciveness to Fusarium wilt. All of the soils were collected from field plots and adjacent areas of the experimental farm at the Central Florida Research and Education Center, Leesburg, FL. All are of the Apopka Fine Sand soil series, which are loamy, siliceous, hyperthermic Grossarenic Paleudults. They have a pH of 6.0-6.5, are low in organic matter (<1%), clay content (<3%), cation exchange capacity (4-6 meq/100 g soil), and available water capacity, and have similar physical and chemical characteristics (Hopkins and Elmstrom, 1984). They differ only in their cropping history and the resulting biology. Soils were collected in large buckets, sieved through a 0.2-cm screen, and stored in plastic bags for up to 4 months before use. The soils are designated as follows.

(1) 'Crimson Sweet' suppressive (CSS), monoculture soil is the Fusarium wiltsuppressive soil developed through monoculture to watermelon cultivar 'Crimson Sweet'. Total populations of F. oxysporum in the soil average 1-2 x 103 cfu/g soil, but the watermelon wilt pathogen apparently accounts for a relatively small proportion of this total








31
(<20%) (Hopkins and Larkin, unpublished). Addition of substantial amounts of the pathogen do not significantly increase disease incidence (Hopkins et al., 1987).

(2) 'Florida Giant' monoculture (FGM) soil is a nonsuppressive soil from the same field as OSS, but it has been monocultured to the susceptible cultivar 'Florida Giant.' High levels of wilt occur in the field, but the soil is similar to OSS in its biology and overall effects due to prolonged monoculture. Total populations of 1-2 x 1 03 cfu/g soil of F. oLxvsgorum are similar to those in OSS soil, with pathogen populations comprising a small, but possibly larger proportion (<30%) of the total population of F. oxvsoorum than in CSS soil (Hopkins and Larkin, unpublished). This soil has been called nonsuppressive rather than conducive because the addition of small amounts of the pathogen cause little change in disease, suggesting this soil does have some biological buffering capacity against the pathogen and is not as conducive as fallow and other soils.

(3) Leesburg fallow, conducive (LFC) soil is from the same vicinity as the monoculture soils, but the area had not been planted to watermelon or any other crop. This represents the natural soil of the area prior to cultivation of watermelon. Total populations of F. oxvsoorum are lower (0.2-1.0 x 10'3 cfu/g soil) than those in the other soils and the pathogen is generally not present; the addition of low levels of the pathogen causes substantial disease.

(4) Microwave-treated 'Crimson Sweet' suppressive (05MW) soil is the CSS soil rendered conducive by microwave-irradiation (2450 MHz, 700 wafts) for 2 mmn/kg soil at a matric potential of -0.01 MPa. This treatment is sufficient to remove all suppressive characteristics and eliminate all F. ox~ysporum and most other fungi, yet leave a large bacterial biomass.








32
Infestation of Soils with Fusarium oxysporum f. sp. niveum and Assay of Fusarium Wilt


Isolates of F. oxysporum f. sp. niveum were obtained from naturally infected watermelon plants in a previous study (Hopkins et al., 1987). For most tests, race 1 isolate FG85-1 was used. Chlamydospores of the pathogen were used as inoculum to simulate natural conditions in the soil. Chlamydospore inoculum was produced in stock soils by the addition of a thoroughly mixed 7-day-old liquid culture, which was grown in a mineral salts solution (Netzer, 1976) and consisted mainly of chopped mycelia and microconidia, to autoclaved (60 min/kg) soil. The soil was moistened, mixed, and allowed to dry. After 4 to 8 weeks, primarily chlamydospores remained and soil dilution plate counts were made to determine their density in soil. A 5-g subsample of soil was added to 45 ml of sterile water and stirred vigorously for 5 minutes. While stirring, 1 ml of this suspension was extracted and added to 9 ml of 0.1% water agar and agitated using a vortex mixer. One milliliter of this suspension, or of a 1 0-fold dilution, was pipetted onto each of four plates of Komada's (1975) selective medium for F. oxysporum and spread over the surface of the agar by tilting the plates. The stock soil (containing approximately lx1 04 chlamydospores/g soil) was then mixed with the soil to be infested to produce the desired inoculum level (generally about 200 chlamydospores/g soil).

Watermelon seeds of susceptible cultivar 'Florida Giant' and moderately resistant cultivar 'Crimson Sweet' were planted in pots of the infested soil (generally four replications of four to six plants in four to six pots varying with individual experiments) in the greenhouse. Plants were maintained at 20-300C and grown for four weeks; maximum light intensity was 500-700 limol/m2/s. Fusarium wilt was assessed by visual inspection of the plants for wilt symptoms several times a week and verified periodically by plating








33

surface-disinfested stem pieces on Komada's (1975) medium. Wilt was expressed as the percentage of diseased plants over the 4-week period. Production and Characterization of Orange Mutant Pathogen Strains


To distinguish F. oxysporum f. sp. niveum added to field soils from indigenous isolates of F. oxvsDorum, an ultraviolet light-induced, orange-colored, mutant strain of the pathogen was produced and characterized for use as a marker organism (Elmer and Lacy, 1987; Puhalla, 1985; Schneider, 1984). Microconidia from 5- to 7-day-old potato dextrose agar (PDA) cultures of a race 1 isolate (FG85-1) of F. oxysporum f. sp. niveum were suspended in water and adjusted to 108 spores/ml. A petri dish containing 10 ml of this suspension was exposed to ultraviolet light (General Electric G8T5 germicidal lamp) for 20 seconds in the dark at a distance of 10 cm; approximately 95% of the spores were killed. Survivors were plated (200-500/plate) on a sorbose-based medium (ingredients/I water: agar, 20 g; sorbose, 20 g; asparagine, 2 g; K2HPO4, 1 g; KCI, 0.5 g; MgSO47H2O, 0.5 g; tergitol NP-1 0, 0.5 ml; and 0.2 ml of a trace element solution composed of citric acid, 5 g; zinc sulfate, 5 g; iron sulfate, 4.75 g; Fe(NH,)2(S04)26H20, 1 g; copper sulfate, 0.25 g; manganese sulfate, 0.05 g; boric acid, 0.05 g; and sodium molybenate, 0.05 g in 95 ml water) and scanned for orange mutants after 6 days. Orange mutants were readily distinguished from the wild-type by their distinct orange pigment.

Growth of the orange mutant (OM) isolates was compared with that of the wildtype parent isolate by measuring radial colony growth and microconidium production on PDA as well as weight increase of mycelia in a liquid medium. Radial growth measurements were made after 3, 5, and 7 days at 260C. After 7 days, 10 ml of sterile water were added to the plates and the surface of the agar gently scraped to loosen the








34
mycelia and conidia. Conidial counts of the resulting suspension were made using a hemacytometer. Mycelial PDA blocks were also transferred to flasks containing 30 ml liquid culture medium (Netzer, 1976) and the flasks incubated at 260C for 5 days. Mycelial mats were harvested on filter paper, washed twice with deionized water and dried at 800C for 24 hr before weighing. Each test was conducted at least twice with four replications per isolate.

Pathogenicity of the OM isolates was tested and compared with that of the wildtype parent isolate in microwave-treated soil and a conducive field soil using both conidia and chlamydospores as inoculum. Conidia were collected from 5- to 7-day-old PDA cultures and added to the soil at the rate of 5 x 103 colony forming units (cfu)/g soil. Chlamydospores were produced as previously described and added to soil to provide 200 cfu/g soil. Watermelon seeds were planted in four to six replicate pots (five seeds per

7.5-cm pot) and wilt was assayed as described above.

Root colonization by the orange mutants was compared with that of the wild-type parent isolate in microwave-treated soil infested with the pathogen. Watermelon seeds were planted (four/pot, four replicate pots/isolate) in soil infested with 5 x 1 03 conidia/g soil or 200 chlamydospores/g soil and grown for 3 weeks. Roots were gently removed from the soil and washed under running water to remove adhering soil particles. Roots were blotted dry, weighed, put in sterile water, and shaken at 150 rpm for 20 minutes. The resulting suspension and a 1:10 dilution were plated on Komada's (1975) medium, and colonies were counted after 5-7 days. Counts were converted to cfu/g root fresh weight. In another technique, roots were washed, separated, cut into sections and plated directly by laying the pieces lengthwise on Komada's (1975) medium, then observed for the number of orange mutant colonies/i 0 cm length of root.








35

Survival and Population Dynamics of Fusarium oxysporum in Field Soils


Chlamydospore inoculum of an OM isolate of the pathogen (FG-OR3) was added to each of the four soils at rates ranging from 2-10 x 102 cfu/g soil, depending on the experiment. Infested soil was maintained either at a constant matric potential of -0.1 or

-0.01 MPa or under a fluctuating moisture regime of alternating wetting and drying cycles. For maintaining a constant matric potential, infested air-dry soil was moistened with deionized water to attain the desired matric potential as determined by soil moisture release curves calculated for each soil. Subsamples of 50 g of each infested soil were kept in small, weighed, plastic screwtop containers. The containers were placed, with tops loosened, in a moist chamber and incubated at 280C. Water was added to the containers weekly to replenish the moisture content and maintain their original weights (variation in matric potential was estimated at t 0.005 MPa). Infested soil samples for the fluctuating moisture regime tests were put in plastic pots in the greenhouse, saturated with water, and allowed to slowly dry. After about 2 weeks, the soil was resaturated and this wetting and drying process was repeated throughout 6 months. Soil samples were taken during the dry phase of the cycle. Periodically, 5-g subsamples were taken from these and used to determine the soil populations of the OM pathogen as well as indigenous F. oxysRorum by soil dilution plating on Komada's (1975) medium as previously described. Populations were monitored over a 6-month period and determined as cfu/g air-dry soil. Tests were made using various initial inoculum levels, and consisted of three replications of each soil, with two samples taken per replication. The fluctuating moisture regime and

-0.1 MPa constant moisture tests were repeated, whereas the -0.01 MPa test was only conducted once.








36
Chlamydospore Germination


Germination of chlamydospores of the pathogen in the four soils was assessed by using the buried membrane filter technique of Adams (1967), with some modifications by Alabouvette et al. (1980, 1985a). Chlamydospores were produced by growing the fungus on carnation leaf agar and suspending the macroconidia in sterile water. The suspension was incubated in the dark at 260C for two weeks, at which time most macroconidia had converted to chlamydospores. Chlamydospore suspensions were adjusted to 2-3 x 104 and 1 ml of this suspension was deposited on a 25 mm millipore membrane filter (type HA 0.45 I.m) with a grid on one side, by vacuum filtration.

A glucose solution was added to 1 00-g subsamples of the test soils at the rates of 0, 0.1, 0.2, 0.4, and 1.0 mg/g soil to give a final matric potential of -0.01 MPa. The treated soils were placed in 5.5-cm-diameter plastic pots and the chlamydosporecontaining membrane filter was buried in the soil by making a slit with a spatula, gently inserting the filter, and covering the filter with soil. The pots were incubated in a moist chamber for 24 hr at 250C. Filters were then gently removed, rinsed in deionized water, stained with trypan blue lactophenol (Adams, 1967), steamed at 700C for 20 min, washed with clear lactophenol by vacuum filtration, and mounted in glycerin on microscope slides. Filters were subdivided into six sectors and at least 100 chlamydospores were counted per sector at a magnification of 200X. Chlamydospores were observed for germination and germ tube length. Germination tests were conducted on representative isolates of race 1 and 2, and tests were repeated at least three times for both isolates.








37

Root Colonization by Fusarium oxysporum


Watermelon cultivars 'Florida Giant' and 'Crimson Sweet' were planted in the four soils infested with the OM pathogen at a rate of about 200 cfu/g soil and grown 2-3 weeks in the greenhouse. Whole root samples were gently removed from the pots and adhering soil removed under running tap water. In initial experiments (surface colonization determinations) roots were not treated further. In later experiments (internal colonization determinations) roots were surface disinfested in 0.5% sodium hypochlorite for 1 min and rinsed in deionized water. All roots were then placed in empty sterile petri plates, covered with sterile water, and the roots separated and teased apart with a dissecting needle. The water was then poured off and the intact root system was embedded in agar by pouring molten Komada's (1975) medium cooled to 450C into the plate covering the entire root. The plates were incubated at 28"C and the colonies emerging from the root were counted 3-7 days later. Colonies of the OM pathogen, indigenous F. oxvsporum, and other fungi could be easily differentiated and quantified. In this way, not only numbers, but the spatial arrangement of the pathogen in relation to other fungi could be observed.

The length of the plated root systems were estimated by a modified line intersect method (Tennant, 1975). The plates containing the roots were placed on a 0.5-cm grid and the number of intersections of a root with a grid line were counted in both the vertical and horizontal directions. Root lengths in cm were determined by multiplication of the total intersections by a conversion factor of 0.393. Root systems ranged from 50-200 cm/plate. Colonization was expressed as number of colonies per 100 cm root. Each experiment was repeated and consisted of four replications of four to six root systems each.








38

Soil and Root Microorganism Populations


Estimates of soil and root microorganism populations were made using standard dilution-plating procedures on various general and selective culture media. Bacterial populations (rapid-growing, aerobic organisms capable of growing readily in culture) were estimated using nutrient agar and 1/10 strength tryptic soy agar. Actinomycetes were selected for on alkaline water agar, pH 10.5 (Ho and Ko, 1980). Populations of fluorescent pseudomonads as well as certain other pseudomonads were determined on selective King's medium B containing penicillin, cyclohexamide, and novobiocin (Sands and Rovira, 1970). Bacterial plates were incubated at 260C for 3-4 days. Plates of King's medium B were examined under ultraviolet light for colonies producing diffusible fluorescent pigments. Actinomycete plates were incubated 7-10 days and total colonies counted. Fungal populations (primarily rapid-growing, spore-forming organisms) were determined on PDA containing 1 ml tergitol NP-1 0 and 50 mg chlortetracycline /I medium. Plates were incubated at 260C for 5-6 days and total colonies counted.

A series of 1:10 dilutions of the initial soil suspensions were made and 0.1-ml aliquots of the appropriate dilutions were spread-plated on the various agar media. Population densities of rhizosphere and rhizoplane microorganisms were estimated from roots of 3-week-old watermelon plants. Roots were gently removed from soil and loosely adhering soil shaken free. Roots were weighed, placed in sterile water in a 1:50 (w/v) dilution, and either shaken on a rotary shaker at 200 rpm for 20 min, or put in a sonicator (Branson ultra-sonic cleaner model B-22-4) for 5 min, depending on the experiment. Appropriate 10-fold dilutions of the resulting suspensions were plated on the various agar media. In some tests, rhizoplane organisms were estimated separately from rhizosphere organisms. After the sonication procedure, the roots were removed from the suspension,








39

rinsed in sterile water, and triturated in a mortar and pestle. Final dilutions were plated as with the others. Populations for all organisms were expressed as log cfu/g soil (Loper et al., 1984). For most experiments, four replications of four plates each were used for each soil or root treatment. All experiments were conducted twice. Selective Elimination of Microorganisms by Microwave Irradiation


The effects on field soils of various exposure times to microwave irradiation on the level of disease suppression (wilt), microorganism populations, and root colonization by F. oxysporum were assessed. Soils were adjusted to a matric potential of -0.01 MPa and microwave-irradiated (2450 MHz, 700 watts) for 0, 30, 60, and 90 s/kg soil. These levels were chosen after preliminary tests (levels ranging from 0-150 s/kg at 15 s intervals) showed disease suppression was eliminated in CSS soil after a microwave exposure of 90 s/kg or longer. Chlamydospore inoculum of the OM pathogen was added to the treated soils at the rate of 200 cfu/g soil. Watermelon cultivars 'Crimson Sweet' and 'Florida Giant' were planted in the soils (four replicate pots of ten plants each for each treatment) and grown for 4 weeks. Wilt was assessed and root microorganism populations and colonization were determined as described above.


Statistical Analyses


All experiments were analyzed using standard analysis of variance (ANOVA) procedures. Significance was evaluated at P<0.05 for all tests and mean separation was accomplished using Duncan's multiple range test. Experimental design for most tests were variations on a randomized complete block. Most simple analyses and individual factor comparisons were conducted using a one way ANOVA. Overall results of two or








40

more factor experiments used a two way ANOVA. Correlation or regression analyses were conducted where appropriate. All data expressed as percentages were arcsi n-transformed (sin" Vx) before analysis. All computations were made using Statgraphics (STSC, Inc.) statistical computer program on a personal computer.


Results


Production and Characterization of Oranaie Mutant Pathogen Strains


A total of eight orange-colored mutant isolates were successfully recovered from parent isolate FG85-1 after repeated screenings with ultraviolet light. The mutation rate was in the range of 10O'-10-'. The eight orange mutant isolates varied considerably regarding the characteristics measured, with some isolates revealing significantly impaired capabilities (reductions of 22-89%) relative to the wild-type parent in radial growth, root colonization, and pathogenicity (Appendix A; Table A-5). Isolates FG-0R3, FG-0R6, and FG-0R8, however, equalled or surpassed FG85-1 for all characteristics measured in initial tests and were found to be indistinguishable from the wild-type in further evaluations for root colonization and pathogenicity (Table 3-1). On the basis of these tests, it was concluded that these isolates could be used to adequately represent F. oxysporum f. sp. niveum in field soils and greenhouse tests. The orange pigment was found to be a stable and reliable marker throughout all phases of this research. Isolate FG-0R3 or FG-0R8 was used in all subsequent experiments involving an orange mutant (OM) pathogen. Survival and Population Dynamics of Fusarium oxysoorum in Field Soils


Differences in the population dynamics of F. oxvsrorum f. sp. niveum were observed among the soils regardless of moisture conditions or initial pathogen inoculum








41

levels (Figure 3-1). At constant matric potentials (-0.1 and -0.01 MPa), populations remained stable, at or near initial inoculum levels (200-600 cfu/g soil), in both monoculture soils (CSS-suppressive and FGM-nonsuppressive) throughout a 6-month study period. In the fallow conducive (LFC) soil, populations increased somewhat within the first few weeks, then remained stable throughout the remainder of the experiment. In the microwave-treated soil (CSMW), there was a dramatic initial increase in propagule numbers followed by a slow decline. Pathogen populations were higher in CSMW soil than in others throughout the study. Populations in the monoculture soils were also lower than in LFC soil at most sampling dates. In soils under a fluctuating moisture regime, represented by alternating wetting and drying cycles, an initial increase in propagules was followed by a gradual decline throughout the six months in all soils. As with constant moisture potential soils, the microwave-treated soil had the greatest increase in propagules and this difference occurred throughout the study period. Chlamvdospore Germination


There were no differences observed for chlamydospore germination in CSS, FGM, or LFC soils at any level of glucose amendments (0-1.0 mg/g soil) (Figure 3-2). Chlamydospores germinated readily (20-80%) in all soils with small additions of glucose (0.1-0.4 mg/g soil). The only difference among soils was at the 0.1 and 0.2 mg glucose/g soil level with isolate CS85-4 and at 0.2 mg/g soil with isolate FG85-1, where CSMW soil resulted in greater germination than any of the field soils. There was some variation between the different pathogen isolates tested at the lower glucose levels, with isolate CS85-4 germinating more readily than FG85-1 at 0.1 and 0.2 mg glucose/g soil, but overall effects were similar. Observational data did not indicate any differences or large








42

variation in germ tube lengths or in the lysis of hyphae between the three field soils at any glucose level.


Root Colonization by Fusarium oxvsporum


Surface colonization levels of 'Crimson Sweet' and 'Florida Giant' watermelon roots by added F. oxysporum f. sp. niveum (OM pathogen) were similar in CSS, FGM, and CSMW soil, whereas colonization was greater in LFC soil than in the other three soils (Table 3-2). Colonization by indigenous F. oxvsporum was similar in CSS, FGM, and LFC soils. There was virtually no colonization in CSMW soil due to the elimination of F. oxysporum by microwave treatment. Colonization by E. oxysporum f. sp. niveum or other F. oxysporum followed no discernible pattern of spatial arrangement on the roots. Colonization by the OM pathogen and other F. oxysporum were intermixed somewhat randomly over the surface. There was no indication of preferential colonization of root tips or young roots; colonization occurred uniformly over all root types. The ratio of colonization by F. oxysporum f. sp. niveum to colonization by F. oxvsporum in CSS soil was not different from that in FGM soil for either cultivar 'Florida Giant' or 'Crimson Sweet,' but the CSS soil ratio was less than for LFC soil for cultivar 'Crimson Sweet.' Level of disease (% wilt), however, was lower in CSS soil (6.6%) than all other soils planted to 'Crimson Sweet' (47-58%). When planted to 'Florida Giant,' CSS soil produced lower wilt (31.8%) than the conducive soils (80.3 and 64.8% for LFC and CSMW soil). The level of Fusarium wilt was not related to the degree of colonization by the OM pathogen, indigenous F. oxysporum, or ratio of OM pathogen/other F. oxysporum. Overall, similar colonization levels were observed on 'Florida Giant' and 'Crimson Sweet' cultivars. Soil populations of the pathogen at the time of root sampling were lower in the monoculture








43

soils than in the conducive soils. With the exception of CSMW soil, which appeared to have unusually low OM colonization in this test compared to subsequent tests, colonization by F. oxysporum f. sp. niveum appeared to be related to F. oxysporum f. sp. niveum populations in the soil rather than to differential colonization in different soils or cultivars.

Since surface colonization appeared to correspond primarily with soil populations, internal colonization (or its approximation) was measured in some tests by plating surface-disinfested watermelon roots. In these tests, in which initial OM inoculum was 200 cfu/g soil, colonization by the OM pathogen averaged 0.91, 1.40, and 6.53 colonies/1 00 cm root for CSS, FGM, and LFC soils, respectively, while the corresponding values for colonization by indigenous F. oxysporum were 10.83, 14.92, and 5.23 colonies/1 00 cm root. There was no difference (Duncan's multiple range test, P<0.05) between CSS and FGM soils for colonization by F. oxysporum f. sp. niveum, F. oxysporum, or in their colonization ratio, although there was a difference between LFC and the two monoculture soils for all three values.


Soil and Root Microorganism Populations


Soil population estimates of general groups of microorganisms in the four soils prior to planting show some overall similarities as well as a few notable differences among the soils (Figure 3-3A). Bacterial populations were similar, ranging from 6.90 to 7.15 log cfu/g soil in the four soils. Estimates of actinomycete populations were significantly greater in the two monoculture soils (6.5 and 6.42 log cfu/g soil, for CSS and FGM, respectively) than in the conducive soils (5.44 and 6.09 log cfu/g soil for LFC and CSMW, respectively). Fluorescent pseudomonad populations were higher in CSS than any other








44

soil. Non-fluorescent pseudomonad-like bacteria populations also were greater in CSS than in FGM and CSMW soils. Fungal populations were highest (and showed the greatest diversity in colony morphology) in LFC soil than any of the other soils (5.42 log cfu/g soil vs. 4.94, 4.75, and 4.48 log cfu/g soil for CSS, FGM, and CSMW soils, respectively). Although actual numbers fluctuated from one test to another, differences among the soil population estimates relative to one another were similar in subsequent tests.

Estimates of microorganism populations isolated from watermelon roots also were different among the four soils (Figure 3-3B). Root populations of bacteria, actinomycetes, fluorescent pseudomonads, and other pseudomonads were greater in both monoculture soils (CSS and FGM) than in LFC or CSMW soil. Fluorescent pseudomonad populations on roots in CSS soil were also greater than in FGM soil (6.00 log cfu/g root in CSS soil and 5.35 log cfu/g root in FGM soil). Fungal populations were greater on roots in LFC than any other soil. Microorganism populations were consistently greater on roots than in soil for all prokaryote groups, with populations averaging 1-2 orders of magnitude greater per gram of roots than per gram of soil. Differences in microorganism populations between soils also tended to be greater on roots than in the bulk soil. Overall fungal populations were comparable both in the soil and on roots, although they averaged slightly lower on roots. All population differences were observed with both 'Florida Giant' and 'Crimson Sweet' cultivars, with no substantial differences observed between cultivars (Appendix A; Table A-6).


Selective Elimination of Microorganisms by Microwave Irradiation


The level of Fusarium wilt observed in soils which had been exposed to microwave treatments of 0 s (control) and 30 s/kg soil, followed by infestation of the pathogen, was








45

significantly different among the three soils (Figure 3-4). These differences reflected the level of suppressiveness of the respective soils; there was less disease in CSS soil than in FGM and LFC soils and FGM soil had less wilt than LFC soil. Microwave exposures of 60 s/kg soil or less had virtually no effect on Fusarium wilt in LFC and FGM soils, while there was a change in the level of suppression in CSS soil. At 90 s/kg soil, all soils showed increased levels of Fusarium wilt, and all differences in suppressiveness between the soils were eliminated. The changes in wilt with increasing microwave exposure (linear regression analysis, P<0.05, b=0.65 0.11 and b=0.34 t 0.11 for CSS and FGM soils, respectively) were different between CSS and FGM soils.

Microwave irradiation at exposures of 60 s or less did not result in any significant change in the internal colonization of roots by f. oxysporum f. sp. niveum (Table 3-3), but the 90-s treatment allowed drastic increases in pathogen colonization in both CSS and FGM soil. Internal root colonization by F. oxvsporum was not different between CSS and FGM soils at any level of microwave exposure. However, microwave treatment of 60 s did reduce Lc<0.05) colonization by f. oxysporum in CSS soil compared to that in untreated soil (5.15 versus 13.94 colonies/100 cm root). No colonization by f. oxvsoorum was observed at the 90-s treatment in either soil. Generally higher levels of colonization by other miscellaneous fungi were observed in FGM than in CSS soil, but microwave exposure of 60 s or less had little effect on colonization. There was little to no colonization by other fungi observed at the 90-s exposure level. Differences between disease levels in the two soils at 0 and 30-s exposures did not appear to be related to colonization either by F. oxvsporum f. sp. niveum or indigenous F. oxysporum, although the reduced level of colonization by indigenous F. oxsporum in CSS soil at the 60-s exposure did coincide with an increased level of wilt observed in that soil.








46

Microwave treatments of up to 90 s/kg soil had no significant effects on root population estimates of total bacteria, actinomycetes, or fluorescent pseudomonads in either soil (Table 3-4). An increase observed in this test in the population of other pseudomonads in both soils at the 90-s treatment was not observed in a repeat of this experiment, in which total root bacterial populations were relatively unaffected by microwave treatment (Appendix A; Table A-7). A decrease in fungal root populations was generally observed at the 90-s treatment, although this effect was more prominent in CSS soil than FGM soil.

Root populations of bacteria, actinomycetes, and fluorescent pseudomonads were generally higher in CSS soil than in FGM soil at the 30-s and 90-s exposure levels (Table 3-4). Root populations of fungi in FGM soil were significantly greater than in CSS soil at 30-, 60-, and 90-s exposure levels. Populations of non-fluorescent pseudomonads were greater in FGM soil at 0- and 60-s exposures. Similar population levels were observed in a repeat of this experiment, although differences between CSS and FGM were not always significant. Combined data from both tests verified the differences in populations of bacteria and fluorescent pseudomonads described here, but actinomycete populations were not different between the two soils (Appendix A; Table A-8).

Rhizoplane organism population estimates demonstrated differences between CSS and FGM soils (Table 3-5). There were some differences in populations due to microwave treatments, but no consistent trends were observed with increasing exposure time. Total bacterial populations were greater on roots in CSS soil at the 0- and 30-s treatments, as were fluorescent pseudomonad populations at the 30-s and 60-s treatments, compared to those in FGM soil. Fungal populations, however, were similar at all microwave treatments except 90 s, where CSS rhizoplane populations were greater than those in








47
FGM soil. There were no differences in other pseudomonad populations at any microwave treatment level (4.0-4.6 log cfu/g soil). Measurements of rhizoplane organism populations were not repeated.

Since, overall, the microwave treatments had a very limited effect on general root microorganism numbers in each soil, the average microorganism populations over all microwave treatments provided an overall comparison between the two soils (Figure 3-5). Average populations of bacteria and fluorescent pseudomonads were greater in CSS soil than in FGM soil for both rhizosphere and rhizoplane organisms, as were actinomycete rhizosphere populations. Fungal rhizosphere populations were greater in CSS soil, but rhizoplane populations were greater in FGM soil.


Discussion


Suppressive soils can be categorized as pathogen-suppressive or diseasesuppressive based on whether they act directly on the pathogen to reduce its population in the soil or act indirectly by reducing the disease-causing activity of the pathogen (Cook and Baker, 1983; Hornby, 1983; Simon and Sivasithamparam, 1989). However, this distinction may not always be known or a soil may incorporate both means of suppression. Components in the soil may suppress the pathogen directly by destruction of propagules or hyphae through lysis or predation, by inhibiting propagule germination, or by reducing saprophytic growth in some other way (Cook and Baker, 1983, Schneider, 1982, Simon and Sivasithamparam, 1989). Suppression of the disease-causing activity of the pathogen may involve reduction in parasitic colonization or infection of the host or reduced ability to induce disease after infection (Cook and Baker, 1983).












Table 3-1. Comparison of selected orange mutant isolates with wild-type parent isolate (FG85-1) of Fusarium oxysporum f. sp. niveum for growth characteristics, root colonization, and pathogenicity.

Radial Conidium Mycelium Root colonizationd PathogqenicityV(% wilt) Isolate growth productionb massc Rootwash Direct MW-treated Field
(mm) (cfu x 106/ml) (mg) (cfu x 103/g) (cfu/10cm) soil soil

FG85-1 38.9' 1.50a 96 3.7 5.9 92 69
FG-OR3 38.0 3.52c 95 4.6 9.7 91 69
FG-OR6 36.9 1.75ab 95 4.3 81 62
FG-OR8 37.9 2.13b 84 8.2 9.9 86 78
a Radial growth was measured on PDA plates after 7 days at 260C (Average of three experiments, each with four replicate plates/isolate).
b Conidium production was estimated from 10 ml conidium suspensions made from 7-day-old PDA cultures (Average of two experiments with four replications/isolate).
SIsolates were grown in liquid medium (Netzer, 1976) for 5 days at 260C. Means represent the dry weight of mycelial mats harvested on filter paper (Average of two experiments with four replications/isolate).
d Weighed root samples (four/isolate) of cultivar 'Florida Giant' were shaken in sterile water 20 minutes and the rootwash suspension dilution-plated on Komada's (1975) medium. For direct-plating, root samples were washed, separated, cut into sections and plated lengthwise. Chlamydospore inoculum of 200 cfu/g soil was used in all tests.
Pathogenicity was measured as the incidence of Fusarium wilt on cultivar 'Florida Giant' in a microwave (MW)treated soil (2 min/kg soil at -0.01 MPa) and a conducive field soil. Chlamydospore inoculum of 200 cfu/g soil was used in all tests (Average of three tests in microwave-treated soil and two tests in field soil using four replicate pots of five plants/test).
f Means within columns not followed by letters or followed by the same letter are not significantly different (jP<0.05) according to Duncan's multiple range test.































Figure 3-1. Population dynamics of Fusarium oxvsporum f. sp. niveum over time in four soils under different moisture regimes (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil; LFC=fallow, conducive soil; CSMW=suppressive soil rendered conducive by microwave treatment). A) Constant matric potential of -0.1 MPa; B) constant matric potential of -0.01 MPa; C) fluctuating moisture potential caused by alternating wetting and drying cycles (soils saturated and allowed to dry in 2-week cycles). Fusarium oxysporum f. sp. niveum was added to soil as an orange-colored mutant isolate. Values within tests at each sampling date denoted by the same letter are not significantly different according to Duncan's multiple range test (P<0o.05).









50






3000 A -0.1 MPa Soil type
C 2500 -CSS + FGM LFC CSMW
f
U
2000
/
1500 o

S 1000
0 b b
S---4 --- b b
500 -- a . . . . . . . . . . . . .,
I o -- - - - --- ------------a
aa a
0
0 1 2 3 4 5 6
Months
3000
000 B -0.01 MPa
Soil type
C 2500 -csS 4-FGM 4-LFC --CSMW
f
u
2000

cc





3500
9 bo b


S 1000 -fa

/ 00
Oaa
500 a aa

0
0 1 2 3 4 5 6
Months

3500 C Wetting/Drying cycles

3000- CSoil type C 3000 -- csMW
f -04- LFC
U 2500- C

2000g 15 0 0bo S


0 1004 a

0


0 1 2 3 4 5 6
Months






























Figure 3-2. Chlamydospore germination of two different isolates of Fusarium oxysporum f. sp. niveum in four soils after additions of glucose (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil; LFC=fallow, conducive soil; CSMW=suppressive soil rendered conducive by microwave treatment). A) Race 1 isolate FG85-1 (average of four tests); B) race 2 isolate CS85-4 (average of three tests). Values denoted by asterisk are significantly different from others at that glucose level within each isolate according to Duncan's multiple range test L.<0.05).








52






100
A

80

G
e
r 60
m

a 40 Soil type
a 40 t --- css
o FGM
n 20 LFC

CSMW

0
0 0.2 0.4 0.6 0.8 1
Glucose level (mg/g soil)

100
B

80

G
e
r 60
m
I1
n
a 40 Soil type
CSS o FGM



0
0 0.2 0.4 0.6 0.8 1
Glucose level (mg/g soil)










Table 3-2. Surface colonization of roots of two different watermelon cultivars by Fusarium oxysporum in four soils in relation to disease incidence and populations of F. oxysporum f. sp. niveum.

Colonies per 100 cm rootb Colonization OM pathogen populationd
Soil type OM pathogen F. oxysporum ratios % Wilt cfu/g soil

'Crimson Sweet'
CSS 6.0ae 27.5b 0.215a 6.6a 44a
FGM 9.8a 33.1 b 0.367ab 51.4b 50a
LFC 26.8b 21.3b 1.360b 58.0b 231 b
CSMW 12.9a 0.2a 10.125c 47.1 b 269b

'Florida Giant'
CSS 8.6a 24.7b 0.518a 31.8a 62a
FGM 11.3a 27.0b 0.429a 43.8ab 38a
LFC 26.4b 17.0b 1.964a 80.3c 450c
CSMW 12.8a 0.3a 19.000b 64.8bc 238b
a Soil type represents differences in the ability of a soil to suppress Fusarium wilt of watermelon. CSS='Crimson Sweet' suppressive, monoculture soil; FGM='Florida Giant' monoculture soil (nonsuppressive); LFC=Leesburg fallow conducive soil; CSMW=microwave-treated, 'Crimson Sweet' soil (conducive).
b Colonization of roots by the orange mutant (OM) strain of F. oxysporum f. sp. niveum and all F. oxysporum other than the OM pathogen was determined by washing the roots of 3-week-old plants and embedding them intact in Komada's (1975) medium. Four replications of four to six roots each were used. Root length was estimated by the line-intersect method (Tennant, 1975).
SThe colonization ratio represents the mean of the colonization by the OM pathogen divided by the colonization by other F. oxysporum calculated for each sample.
d Soil population of OM pathogen was determined by dilution-plating at the time of root colonization measurements. Initial inoculum was approximately 100 cfu/g soil.
Means within columns for each cultivar followed by the same letter are not significantly different (P<0.05) according to Duncan's multiple range test.































Figure 3-3. Population estimates of bacteria, actinomycetes, fluorescent pseudomonads, other pseudomonads, and fungi in soil and on watermelon roots in four different soils (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil; LFC=fallow, conducive soil; CSMW= suppressive soil rendered conducive by microwave treatment). A) Soil microorganism populations prior to planting to watermelon; B) microorganism populations on 3-week-old watermelon roots (combined values for cultivars 'Florida Giant' and 'Crimson Sweet'). Values within each microorganism group topped by the same letter are not significantly different LP<0.05) according to Duncan's multiple range test.








55






10
A
A Soil type

L css
o 8g b FGM
C C LFC
f 6- CSM

a
/ a


g 4 b a
S

2
II

0
bacteria actinomycetes fluor. pseudo. other pseudo. fungi Microorganism group

10
Soil type L css
o 8 g W FGM
C LFC
f C CSMW
U


g 4
r

t


0
bacteria actinomycetes fluor. pseudo. other pseudo. fungi Microorganism group









56




P r C
e a
n 100
t

S80

t 60 b b


40 FC

20 a FGM

0 -CSS
0 30 60 90
Microwave exposure (s)






Figure 3-4. Fusarium wilt development in three soils exposed to varying length microwave exposures (CSS =suppressive, monoculture soil; FGM= nonsuppressive, monoculture soil; LFC=fallow, conducive soil). Chlamydospore inoculum of 200 cfu/g soil was added to each soil following microwave treatment. Values within a given exposure level topped by the same letter are not significantly different (P<0.05) according to Duncan's multiple range test.








57

Table 3-3. Internal colonization of 'Crimson Sweet' watermelon roots by Fusarium oxvsporum in two soils exposed to varying microwave treatments.

Microwave Colonies per 100 cm root
Soil type exposureb OM pathogens F. oxysporumd Other fungie % Wilt

CSS 0 1.02a 13.94b 1.37a 6.2a
FGM 0 0.55a 14.34b 5.77b 35.5bc

CSS 30 0.80a 7.72ab 1.66a 14.Sab
FGM 30 2.24a 15.51 b 4.11 ab 37.5c

CSS 60 0.71 a 5.15a 3.26a 33.7bc
FGM 60 0.34a 8.11 ab 11.11b 43.Oc
CSS 90 7.48b 0.00a 0.63a 78.7d
FGM 90 14.19c 0.00a 0.00a 82.5d
a Soil type represents differences in the ability of a soil to suppress Fusarium wilt of watermelon. CSS='Crimson Sweet' suppressive, monoculture soil; FGM ='Florida Giant' monoculture soil (nonsuppressive).
b Duration of microwave exposure in s/ kg soil (2450 MHz, 700 watts) at -0.01 MPa matric potential. Following microwave treatment all soils were infested with the OM pathogen at 200 cfu/g soil.
Colonization of roots by an orange mutant (OM) strain of F. oxysporum f. sp. niveum (FGOR-3) was determined on the roots of 3-week-old plants that had been washed, surface-sterilized in 0.5% sodium hypochlorite for 1 min, rinsed, and embedded intact in Komada's (1975) medium. Four replications of four to six roots each were used. Root length was estimated by line intersect method (Tennant, 1975).
d All F. oxvsporum other than the OM pathogen was counted; roots were prepared as described above.
e All other fungi which were capable of restricted growth on Komada's (1975) medium were counted.
Means in columns followed by the same letter are not significantly different (p<0.05) according to Duncan's multiple range test. Additional statistical comparisons (ANOVA, Duncan's multiple range test) between the two soils at each microwave exposure level as well as within each soil over microwave exposures did not result in any significant differences other than those already indicated. Linear regression analysis of wilt data (.<0.05) also demonstrated a difference between the two soils with increasing microwave exposure (b=0.65 0.11 for CSS soil and b=0.34 -_ 0.11 for FGM soil).








Table 3-4. Microorganism populations on 'Crimson Sweet' watermelon roots in two soils exposed to varying microwave treatments.

Microwave Log cfu / q rootc
Soil type' exposureb Bacteria Actinomycetes Fluorescent Other Fungi
(s/kg soil) pseudomonads pseudomonads

CSS 0 8.03abd 7.25ab 5.51 bc 5.68a 4.40bc
FGM 0 7.68a 7.10Oa 5.18ab 6.23b 4.55bcd

CSS 30 8.48c 7.58c 5.54bc 6.07ab 4.26abc
FGM 30 8.03ab 7.17ab 5.04a 6.33bc 5.10d

CSS 60 7.87ab 7.29ab 5.19ab 5.70a 4.01 ab
FGM 60 7.78ab 7.1 Oa 4.51 a 6.22b 5.02d
CSS 90 8.17bc 7.33b 6.24c 6.64c 3.66a
FGM 90 7.86ab 7.10Oa 5.16ab 6.64c 4.74cd
a Soil type represents differences in the ability of a soil to suppress Fusarium wilt of watermelon. CSS='Crimson Sweet' suppressive, monoculture soil; FGM='Florida Giant' monoculture soil (nonsuppressive).
b Values represent the duration of microwave exposure (s/kg soil) (2450 MHz, 700 watts) at -0.01 MPa matric potential. Following microwave treatment all soils were infested with OM pathogen at 200 cfu/g soil.
c Estimates of microorganism populations were made by sonication of roots from 3-week-old plants in sterile water for 5 min and the resulting suspensions dilution-plated on various agar media. Four replications of two roots each were used. Bacterial populations were estimated on nutrient agar and 1/10 strength tryptic soy agar; actinomycete populations were estimated on alkaline water agar; pseudomonad populations were estimated on a selective King's medium B (Sands and Rovira, 1970). Fluorescent strains were identified by the production of diffusible fluorescent pigment when plates were examined under UV light. Fungal populations were estimated on potato dextrose agar with tergitol and chlortetracycline added.
d Means in columns followed by the same letter are not significantly different @<0.05) according to Duncan's multiple range test. Additional statistical comparisons (ANOVA, Duncan's multiple range test) between the two soils at each microwave exposure level as well as within each soil over microwave exposures did not result in any significant differences other than those already indicated. Linear regression analysis over microwave exposure was not significant (E<0.05) except for populations of fluorescent pseudomonads, other pseudomonads, and fungi in CSS soil (b=0.007 0.002, -0.016 0.003, and -0.0065 0.0025, respectively) and fluorescent pseudomonads in FGM soil (b=0.0073 0.0019).
01
co








59

Table 3-5. Rhizoplane organism populations in two soils exposed to varying microwave treatments.

Microwave Log cfu / ,g root
Soil type exposure Bacteria' Fluorescent Fungi'
(s/kg soil) pseudomonadsd

CSS 0 5.95bc' 4.37b 2.59abc
FGM 0 5.48a 3.40ab 2.74bcd

CSS 30 5.94bc 4.04b 2.07a
FGM 30 5.48a 2.77a 2.12a

CSS 60 6.04c 4.16b 3.22d
FGM 60 5.78abc 2.82a 2.79cd
CSS 90 5.92bc 3.77ab 3.1Ocd
FGM 90 5.69ab 3.69ab 2.15ab
a Soil type represents differences in the ability of a soil to suppress Fusarium wilt of watermelon. CSS ='Crimson Sweet' suppressive, monoculture soil; FGM ='Florida Giant' monoculture soil (nonsuppressive).
b Values represent the duration of microwave exposure (s/kg soil) (2450 MHz, 700 watts) at -0.01 MPa matric potential. Following microwave treatment all soils were infested with OM pathogen at 200 cfu/g soil.
0 Estimates of rhizoplane bacterial populations were made after sonication of roots from 3-week-old plants in sterile water for 5 min. Roots were rinsed in sterile water and triturated in sterile water with a mortar and pestle. The resulting suspensions were dilution-plated on nutrient agar and 1/10 strength tryptic soy agar. Four replications of two roots each were used.
d Fluorescent pseudomonads were estimated on King's medium B with penicillin, cyclohexamide, and novobiocin added. Plates were examined under UV light for the production of diffusible fluorescent pigment.
Fungal populations were estimated on potato dextrose agar with tergitol and chlortetracycline added.
Means in columns followed by the same letter are not significantly different (P<0.05) according to Duncan's multiple range test. Additional statistical comparisons (ANOVA, Duncan's multiple range test) between the two soils at each microwave exposure level did not result in any significant differences other than those already indicated. Comparisons within CSS soil over microwave exposure did indicate some additional significant differences in fluorescent pseudomonad and fungal populations among microwave exposures. Linear regression analysis over microwave exposure was not significant (P<0.05) except for populations of fluorescent pseudomonads and fungi in CSS soil (b=-0.0055 t 0.0015 and 0.0089 t 0.0035, respectively) and bacteria in FGM soil (b=0.0030 0.0013).






























Figure 3-5. Rhizosphere and rhizoplane microorganism populations on 3-week-old 'Crimson Sweet' watermelon roots in two soils (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil). A) Rhizosphere population estimates; B) rhizoplane population estimates (Averages over all microwave treatments). Values within each microorganism group topped by the same letter are not significantly different (~<0.05) according to Duncan's multiple range test.









61







10o A Rhizosphere
Soil type
L CSS
o
g 8- FGM
c
f
u 6


g 4r
0
o 2
t


0
bacteria actinomycetes fluor. pseudomonads fungi
Microorganism group


7 B Rhizoplane Soil type

Lo 6 l css
g FGM
c 5
f
U
4

g 3
a
r
o 2
0
t


0
bacteria fluor. pseudomonads fungi
Microorganism group








62

In this study, there was no evidence of active destruction of pathogen propagules in the suppressive soil. Populations of Fusarium oxvsporum f. sp. niveum were stable over a 6-month period, with no substantial decrease in propagules over time. While the pathogen did multiply somewhat in the conducive soils (initial increases in field soil as well as microwave-treated soil), it did not multiply in the monoculture soils. This indicates some differences in fungistasis among the soils, since the ability of an organism to multiply in a soil is one of the criteria used to determine the presence or absence as well as level of microbiostasis (Ho and Ko, 1982).

Chlamydospore germination has been used as an indicator of the fungistatic activity and suppressiveness of soils (Cook and Baker, 1983; Huang et al., 1988; Hwang et al., 1982). No differences were observed in pathogen chlamydospore germination as a response to glucose amendments among the three field soils used in this study. There were no differences noted in germ tube lengths or in the lysis of hyphae. Thus, there was no indication of suppression or reduction of pathogen saprophytic growth within the suppressive soil by these measures. Higher germination was noted in CSMW soil at some glucose levels, but this soil had its microbiology, nutritional status, and chemistry seriously altered by microwave treatment, and would be expected to have had a low level of fungistasis.

No relationship was observed between suppressiveness and chlamydospore germination in these soils. This differs distinctly from most other Fusarium wilt-suppressive soils that have been studied, in which suppressiveness has been consistently associated with a high level of fungistasis, reduction of saprophytic growth, and the inhibition of chlamydospore germination (Alabouvette, 1986; Alabouvette et al., 1985b; Louvet et al. 1981; Cook and Baker, 1983; Schneider, 1982; Smith, 1977). Compared to








63
chlamydospore germination studies in other Fusarium wilt-suppressive soils, CSS soil required much lower levels of glucose to overcome fungistasis and initiate germination. At 0.4 mg glucose/g soil, chlamydospore germination in CSS soil was around 75-80%; whereas in several other reported suppressive soils, chlamydospore germination averaged only 10-35% at similar glucose levels (Alabouvette et al. 1980, 1985a; Hwang et al., 1982; Smith and Snyder, 1972; Sneh et al., 1984). In these other studies, germination rates in conducive soils averaged from 60-90% at around 0.5 mg glucose/g soil, indicating that fungistasis levels in soils in this study (CSS, FGM, and LFC) are uniformly low and comparable to conducive soils in other studies. Since suppression related to inhibition of chlamydospore germination is usually attributed to a general nutrient competition due to a large antagonistic microbial biomass, it appears that this general suppression evident in other suppressive soils is probably not a major factor in the CSS suppressive soil. However, since the suppressive soil did maintain lower pathogen populations and the pathogen was unable to multiply as readily as in conducive soils, saprophytic growth and development may be affected at some stage other than chlamydospore germination.

High populations of nonpathogenic F. oxysporum have been shown to be involved in suppressiveness in some Fusarium wilt-suppressive soils. Alabouvette and others (Alabouvette, 1986; Alabouvette et al., 1985b, Louvet et al. 1981) determined that the primary cause of suppression in the Chateaurenard region soils is a result of intrageneric competition in the immediate vicinity of the roots during the saprophytic development that precedes establishment of F. oxvsporum at the root surface. Schneider (1984), who associated nonpathogenic F. oxvsporum with suppression of celery wilt in a California soil, attributed the mechanism of suppression to parasitic competition for infection sites








64
on the root. Thus, he distinguished the saprophytic competition that precedes root colonization from parasitic competition for root infection. Tamietti and Pramotton (1990) attributed the suppression of flax wilt in three soils in Italy to nonpathogenic strains of F. oxysporum by a similar mechanism. Paulitz and co-workers (1987, Park et al., 1988) introduced isolates of F. oxysporum which reduced Fusarium wilt of cucumber in field soils, but the mechanism was not determined. Pathogenic F. oxysporum are known to be capable of superficially infecting and colonizing a wide number of nonsusceptible hosts without causing disease (Armstrong and Armstrong, 1948; Banihashemi and deZeeuw, 1975; Hendrix and Nielsen, 1958; Katan, 1971), and may be nearly as effective as nonpathogens in colonizing nonsuscepts and crop residues (Elmer and Lacy, 1987; Gordon et al., 1989; Gordon and Okamato, 1990). Many of the so-called "nonpathogens" in these studies may actually be pathogens of other hosts.

Other studies have shown that nonpathogenic or avirulent F. oxysporum strains applied to roots can protect the host from disease when challenged by a virulent strain (Biles and Martyn, 1989; Davis, 1967, 1968; Gessler and Kuc, 1982; Martyn et al. 1990; Ogawa and Komada, 1984, 1985, 1986; Shimotsuma et al., 1972; Wymore and Baker, 1982). Ogawa and Komada (1986) demonstrated this protection to be a result of an induced systemic resistance caused by previous infection by F. oxysporum. Although induced resistance was also implied or suggested in many other studies, it has generally not yet been proven to be the mechanism involved. In most of these studies, cut, wounded, or bare roots were dipped in concentrated conidial suspensions of the antagonist and then challenged with the pathogen in the same way with these same roots. Thus, the possibility remains that some type of competition for infection may be occurring on or within the root. Moreover, an induced resistance of the type described








65
has not yet been shown to occur under natural conditions and it is not known whether such a mechanism may be operating in suppressive soils (Louvet, 1989; Matta, 1989).

In this study, surface colonization and internal colonization of watermelon roots by F. oxvsporum were analyzed, so that the possibility of both saprophytic and parasitic competition could be evaluated in this suppressive soil. Overall, colonization levels were similar to those reported for F. oxvsporum by others (Gordon et al., 1989; Gordon and Okamato, 1990). However, neither surface nor internal colonization measurements revealed significant differences between CSS and FGM soils, although there were differences between the two monoculture soils and the two conducive soils. Surface colonization levels appeared to be affected primarily by soil population levels. Internal colonization by E. oxvsrorum f. sp. niveum averaged consistently lower in 055 soil, but was not significantly different than colonization in FGM soil. Differences in wilt among the soils did not appear to be related to the levels of colonization by the OM pathogen, indigenous F. oxvsporum, or the ratio of OM pathogen/other F. oxvslorum. Thus, saprophytic competition provided by large populations of nonpathogenic F. oxvsiorum, as observed by Alabouvette (1986), is apparently not the mechanism responsible for suppression in this soil.

Since colonization by the general population of F. oLxvsporum was not related to suppression, it could indicate that specific strains of F. oxvysporum rather than general population levels may be responsible. Schneider (1984) observed that many strains of F. oxvsrorum were capable of infecting roots, but only some of these were effective antagonists. This may explain why the level of colonization by F. oxvsporum is comparable in 055 and FGM soils, yet only CSS is suppressive. The suppressive soil may contain more isolates capable of reducing disease, which are enhanced by the








66
cultivar 'Crimson Sweet.' These strains may not colonize roots more effectively than others, but may be more effective in suppressing disease.

Indigenous isolates of F. oxysporum f. sp. niveum which were not marked with the orange phenotype may be present in the field soils used in these experiments. Since these would be measured as indigenous F. oxvsiorum, they must be taken into consideration, particularly if the numbers of indigenous pathogens differ in CSS and FGM soils. Wilt of susceptible cultivars in unamended CSS soil is consistently low (0-20%), whereas it can be quite variable in FGM soil, ranging from 10-70% wilt and averaging about 30-50% wilt in greenhouse trials (unpublished). However, disease levels in either soil often do not change substantially with addition of moderate amounts of the pathogen (up to 300 clamydospores/g soil); differences in disease levels between the soils may not necessarily be caused by different pathogen populations. Based on pathogenicity tests of limited numbers of isolates of F. oxvsporum taken from the soil and roots, the proportion of indigenous pathogenic isolates of F. oxysporum f. sp. niveum relative to the total population of F. oxysporum in CSS soil is thought to be quite small (<20%), and pathogen populations in FGM soil appear to be only slightly higher (<30%) (Hopkins and Larkin, unpublished). Although it is not known what percentage of root colonization by indigenous F. oxvsporum is actually due to unmarked isolates of the pathogen in these experiments, plating of surface-disinfested stems and roots from diseased plants only rarely revealed systemic infection by non-orange-mutant pathogens. Moreover, wilt in the root colonization experiments resulted from systemic infection by the OM pathogen only, so relationships between colonization and level of wilt were not affected. Thus, although a slightly larger percentage of colonization by F. oxysporum in FGM soil may have been from pathogenic strains, the effect on the overall results appears to be minimal.








67

Nonetheless, the possibility of parasitic competition on the root cannot be fully evaluated without additional experiments that can more accurately monitor colonization by pathogens versus nonpathogens and comparisons of specific antagonist strains.

Population estimates of total bacteria, actinomycetes, fluorescent pseudomonads, and fungi in the soil and on root surfaces were similar to those reported by others using similar soils and methodology (English and Mitchell, 1988; Simon and Sivasithamparam, 1988). Populations of these microorganism groups showed significant differences among the different soils, with fluorescent pseudomonad populations on watermelon roots and in the soil being consistently greater in CSS soil than all others. 'Crimson Sweet' suppressive soil also supported significantly greater bacteria and actinomycete root populations than the other soils in some tests. The substantially higher microorganism populations found on roots as opposed to within the soil in all soils demonstrate the importance of the root on soil microorganisms (Foster and Bowen, 1982). The greater effect of the root on bacterial populations over fungi is a result of the greater ability of bacteria to quickly colonize roots and most effectively utilize root exudates. Fluorescent pseudomonad populations showed the largest increases throughout this study, both due to soil differences and due to root influence. Fluorescent pseudomonads are known to be very effective and competitive root colonizers and have often been associated with disease suppression and promotion of plant growth (Schroth and Hancock, 1982). The differences in fluorescent pseudomonad populations between OSS and FGM soils suggest that fluorescent pseudomonads may be responding to differences in root exudates between cultivars, and thus may be important in the suppressive response.

Exposure of the soils to microwave treatments had little effect on overall microorganism numbers on watermelon roots, and there was also no difference in the








68

relative effect between CSS and FGM soils. This indicated that although there were differences in population levels between the soils, there was no obvious differential response due to composition of organisms which may have responded differently to heat and microwave treatment. However, these experiments did demonstrate that at the point at which populations of indigenous F. oxvsporum were eliminated (90 s), suppression was lost, and where populations of F. oxysporum were suppressed markedly (60 s), disease suppression was also partially lost. At the same time, microwave exposure had no significant effect on overall numbers of bacteria, actinomycetes, or pseudomonads on watermelon roots. Overall fungal populations were also not as severely affected as F. oxvsporum at the 90-s treatment. However, although overall numbers may not have been affected, species composition and diversity, which were not monitored in this study, may have been drastically altered by microwave exposure. Microwave treatments are known to have a greater effect on fungal populations than on bacterial populations, and are considered more desirable than autoclaving or fumigation for eliminating fungal pathogens because it is less disruptive and leaves a large, relatively diverse bacterial biomass (Ferriss, 1984). This is indirect evidence that F. oxysrorum or some other fungi, rather than bacteria, may be important in the suppressive response.

Throughout these tests little difference has been noted between OSS and FGM soils, even though there is a large difference between these soils in the level of wilt observed in the field. When compared to conducive soils, many of the characteristics of these soils appear to be the result of monoculture and not specific differences related to cultivars, yet only CSS soil is actually suppressive, whereas monoculture soils from 'Florida Giant' and other cultivars, although not suppressive, are not as conducive as








69
fallow or rotation soil (Hopkins et al., 1987). Determining the significant difference between these soils is critical to understanding this suppression.

The suppressive soil used in this study had already been shown to be different from the majority of the described Fusarium wilt-suppressive soils in many ways (Hopkins et al., 1987). In the current study, many additional characteristics of the ecology of the pathogen in this soil have been analyzed, indicating additional differences from other Fusarium wilt-suppressive soils, as well as possible interactions and mechanisms important in the suppressive response. This soil appears to be disease-suppressive, rather than pathogen -suppressive, in that the disease-causing activity of infection and development of disease is where suppression most probably occurs. From this work, the organisms showing the largest differences between suppressive and conducive soils, and thus the most important organisms to study further for their interactions with the pathogen, were indigenous F. oxysporum and fluorescent pseudomonads. It is most interesting that these are the two groups most often associated with antagonism or suppression of Fusarium wilt in other soils (Cook and Baker, 1983; Alabouvette, 1986; Alabouvette et al., 1 985b; Huang et al., 1988; Louvet et al., 1981; Scher and Baker, 1980, 1982). However, other organisms not yet identified from the general microorganism groups monitored in this study may also be important in suppression.

This study has dealt with general ecological characteristics and general soil and root populations of microorganisms. Cultivation to watermelon appeared to have some general effects on the microorganism populations in these soils. The key to understanding this suppression may be in specific antagonistic isolates which are different among the soils. Cultivar 'Crimson Sweet' may promote specific strains with characteristics which make them more effective as antagonists and are present in greater








70

numbers in CSS soil. The next step in this research will be to analyze the specific effects of successive planting of 'Crimson Sweet' on microorganism populations followed by a close look at specific antagonists and their interactions with the pathogen.














CHAPTER 4
THE EFFECT OF SUCCESSIVE WATERMELON PLANTINGS ON
FUSARIUM OXYSPORUM AND OTHER MICROORGANISMS IN
SOILS SUPPRESSIVE AND CONDUCIVE TO FUSARIUM WILT OF WATERMELON Introduction


Soil suppressiveness to plant disease can occur naturally (inherent within the chemistry and biology of the soil and independent of cropping) or be induced by some practice or activity, such as planting a crop, the addition of organisms or nutritional amendments, or a particular cultural practice, that causes a change in the microflora environment (Baker and Cook, 1974; Cook and Baker, 1983; Hornby, 1983; Shipton, 1977). Induced suppressive soils are exemplified by the occurrence of take-all decline of wheat, in which suppressiveness to Gaumannomyces araminis var. tritici results after several years of continuous monoculture to wheat (Cook, 1981; Gerlagh, 1968; Shipton, 1975). A similar induction of suppressiveness has been observed over a much shorter time period for Rhizoctonia solani on radishes, alfalfa, and sugar beets (Chet and Baker, 1980; Henis et al., 1978, 1979). In these soils, suppressiveness develops as a result of the build-up of antagonists in response to high pathogen populations produced by the successive growing of susceptible cultivars. Suppression of take-all has been associated with certain fluorescent pseudomonads as well as other bacteria (Cook and Rovira, 1976; Weller and Cook, 1983; Weller et al., 1988), while the suppression of Rhizoctonia solani is attributed to Trichoderma spp. (Chet and Baker, 1980; Henis et al., 1978, 1979; Liu and Baker, 1980).

71








72
Most soils known to be suppressive to Fusarium wilt diseases are naturally occurring (Alabouvette, 1986; Alabouvette et al., 1985b; Cook and Baker, 1983; Louvet et al., 1981; Tamietti and Pramotton, 1990). However, a soil suppressive to Fusarium wilt of watermelon that was induced by monoculture of a particular cultivar of watermelon ('Crimson Sweet') has been described and studied (Hopkins et al., 1987; Chapter 3). Few other examples of induced wilt-suppressive soils have been reported. Sneh et al. (1987) reported a soil suppressive to Fusarium oxvsporum f. sp. melonis that was induced by continuously cropping to resistant melon varieties for several years. Evidence of a similar induction of suppression in the early 1900's was recounted by Kommedahl et al. (1970), in which long-term monoculture of a cultivar resistant to flax wilt resulted in a marked decline in disease following several years of increases, whereas cropping to susceptible cultivars resulted in complete wilt (100%) every year. Schneider (1984), also observed what may have been an induced suppression to Fusarium wilt of celery where "islands" of healthy celery plants were found in fields otherwise uniformly devastated by wilt. In both of these most recent cases, the organisms responsible were concluded to be isolates of F. oxysporum not pathogenic to the crop plant.

Previously, the population dynamics and chlamydospore germination of F. oxysporum f. sp. niveum, as well as root colonization by F. oxysporum and other microorganism groups, was studied in the 'Crimson Sweet' suppressive monoculture soil and compared with similar conducive soils; some distinct differences were demonstrated among these soils (Chapter 3). Because the induction of suppression in this monoculture soil is linked with cultivation of a particular cultivar, it is valuable to study the effects of such cultivation on different microorganism groups. By comparing the changes in soil and rhizosphere microflora populations directly due to the planting of watermelon and the








73
differences between cultivars, the important microbiological interactions and their role in suppression can be determined more accurately.

Because previous studies with this suppressive soil, as well as other Fusarium wiltsuppressive soils, have indicated an important role for nonpathogenic E. oxvsporum, special emphasis was placed on changes in the populations of F. oxysporum and watermelon root colonization by F. oxysporum. Using four soils representing different suppressive and conducive conditions, the objectives of this study were to evaluate the effect of successive plantings of different watermelon cultivars on the population dynamics and root colonization of F. oxysporum f. sp. niveum, indigenous F. oxysporum, and other general microorganism groups, and the incidence of Fusarium wilt.


Materials and Methods


Soils Used


The four soils used throughout this study to represent different suppressive and conducive conditions to Fusarium wilt of watermelon are from the Central Florida Research and Education Center, Leesburg, FL, and have been described previously (Chapter 3). All are of the Apopka Fine Sand soil series (loamy, siliceous, hyperthermic Grossarenic Paleudults) and have similar physical and chemical characteristics. They differ primarily in their cropping history and the resulting biology. The soil designations and suppressiveness rankings are as follows: 'Crimson Sweet' monoculture soil (CSS) (suppressive); 'Florida Giant' monoculture soil (FGM) (nonsuppressive); Leesburg fallow soil (LFC) (conducive); and microwave-treated CSS soil (CSMW) (conducive).








74
Successive Plantings of Watermelon and Assay of Fusarium Wilt


An orange-colored mutant strain of the pathogen (FG-OR3), which was comparable to the wild-type, race 1 strain in growth, pathogenicity, and root colonization, was used to distinguish the pathogen from indigenous F. oxysporum in the field soils (Chapter 3). Soils were infested with chlamydospore inoculum of the orange mutant (OM) pathogen as was described previously (Chapter 3).

Infested soil was put in plastic pots in the greenhouse and allowed to equilibrate for approximately 3 weeks. Watermelon seeds of 'Florida Giant,' a cultivar susceptible to Fusarium wilt, or 'Crimson Sweet,' a moderately resistant cultivar associated with soil suppressiveness, were planted in the infested soil (10 seeds/pot, four to six replicate pots/treatment). Plants were maintained at 20-300C with a maximum light intensity of 500700 imol/m2/s, and were grown for 4 weeks. Fusarium wilt was assessed by visual inspection of the plants for wilt symptoms several times a week and verified periodically by plating surface-disinfested stem pieces on Komada's (1975) selective medium for F. oxysporum. Wilt was expressed as the percentage of diseased plants over the 4-week period. After the final wilt assessment, plants were harvested, soil and root samples were collected, the soil was mixed thoroughly within the pot, and the pots were replanted with the same cultivar in the same manner. This was continued for four to five successive plantings over a 6-month period. All tests were conducted at least twice. Population Dynamics and Root Colonization of Fusarium oxysporum


Soil samples of 5 g each were taken from each pot at the time of initial infestation and immediately prior to each successive planting. Populations of F. oxysporum f. sp. niveum (as represented by the OM pathogen) and indigenous F. oxysporum were








75

determined by serial dilution plating on Komada's (1975) medium as described previously (Chapter 3).

Whole root samples of 3-week-old watermelon plants were gently removed from the pots and rinsed under running water. In some experiments, surface colonization was determined by embedding the roots in molten Komada's (1975) medium as previously described (Chapter 3). Internal colonization was determined by surface-sterilizing roots in 0.5% sodium hypochlorite for one minute, rinsing in sterile water, then plating out as with the others.

Colonies of the OM pathogen, other F. oxysporum, and other fungi were differentiated by color and morphology. In addition, the spatial arrangement of the pathogen in relation to indigenous F. oxysporum and other fungi was observed. The lengths of plated root systems were estimated by a line intersect method (Tennant, 1975) and colonization was expressed as colonies/1 00 cm root. All experiments used four replications of four to six root systems each and were conducted twice (Chapter 3).


Soil Microorganism Populations


Estimates of general soil microorganism populations were made using standard serial dilution plating procedures as described previously (Chapter 3). Overall populations of aerobic, heterotrophic bacteria were estimated using nutrient agar and actinomycete populations were estimated on alkaline water agar, pH 10.5 (Ho and Ko, 1985). Fluorescent pseudomonad populations were estimated using selective King's medium B with cyclohexamide, penicillin, and novobiocin added (Sands and Rovira, 1970). General fungal populations were estimated on potato dextrose agar with 1 ml tergitol NP-10 and 50 mg chlortetracycline added per liter. All plates were incubated at 260C. Nutrient agar








76
and King's medium B plates were incubated 3-4 days, and King' medium B plates were examined under ultraviolet light for colonies producing diffusible fluorescent pigments. Alkaline water agar plates were incubated 7-10 days and total colonies counted. Fungal plates were incubated 5-6 days. Populations were expressed as log colony forming units (log cfu)/g soil and four replications of four plates/treatment were- used.


Statistical Analyses


All experiments were analyzed using standard analysis of variance (ANOVA) procedures. Significance was evaluated at P

Results


Population Dynamics of Fusarium onvsporum and Fusarium Wilt


Five successive plantings of watermelon cultivars 'Florida Giant' and 'Crimson Sweet' resulted in distinct differences in the population of F. oxvYsoorum f. sp. niveum among the four soils (Figure 4-1). After the equilibration period (3 weeks) and at the time of the first planting, pathogen populations in the suppressive (CSS) soil were lower than








77

in the two conducive soils (LEO and CSMVW) for both cultivars, but were not different than those in nonsuppressive monoculture (FGM) soil. This increase in the pathogen populations of the conducive soils before planting was similar to that observed previously in unplanted soils (Chapter 3). Significant differences were noted in OM pathogen populations among the conducive soils and the monoculture soils beginning with the second planting of 'Florida Giant' and at each subsequent planting. With cultivar 'Crimson Sweet,' pathogen populations in the two monoculture soils were also lower than in the conducive soils at all plantings except the fourth.

Comparisons within each soil (ANOVA, P<0.05) determined that successive plantings of either cultivar 'Florida Giant' or 'Crimson Sweet' had no effect on pathogen populations in either monoculture (CSS or FGM) soil throughout the study period. In the conducive soils, plantings of 'Florida Giant' resulted in increasing pathogen populations with each successive planting (ANOVA, linear regression analysis, P<0.05, b=3.31 t .41 and 3.05 t .29 for CSMW and LEO soil, respectively). With planting to 'Crimson Sweet,' however, after the initial population increases in the raw soils, pathogen populations levelled off and stabilized, showing no further increases throughout the remaining successive plantings in the conducive soils.

Thus, there were also differences observed in comparisons between CM pathogen populations in the conducive soils planted to 'Florida Giant' versus 'Crimson Sweet;' CSMW and LEO soil planted to 'Florida Giant' averaged 1956 and 1613 cfu/g soil by the fifth planting, while CSMW and ILEO soil planted to 'Crimson Sweet' averaged 881 and 731 cfu/g soil, respectively. Populations of the pathogen were more stable in the monoculture soils; by the fifth planting, OSS and EGM soils had averages of 206 and 250








78
cfu/g soil, respectively, when planted to 'Crimson Sweet' and 375 and 393 cfu/g soil with 'Florida Giant.'

Estimates of soil populations of indigenous F. oxysporum (all F. oxysporum other than the OM pathogen) also demonstrated differences with successive plantings of 'Florida Giant' and 'Crimson Sweet' (Figure 4-2). When planted to 'Florida Giant,' populations of F. oxysporum fluctuated widely in each soil, but showed no consistent change with successive plantings and stayed within the normal range associated with that soil. Soil populations of F. oxysporum also fluctuated when planted to 'Crimson Sweet,' but showed a significant trend of increase with successive plantings in LFC, FGM, and CSS soils (ANOVA, linear regression analysis, P<0.05, b=3.18 .48, 2.23 .67, 3.88 _ 1.67, for LFC, FGM, and CSS soils respectively). By the fifth planting of 'Crimson Sweet,' populations averaged 2630, 2220, and 1593 cfu/g soil in CSS, FGM, and LFC soils, respectively, as compared to 1098, 1424, and 212 cfu/g soil, respectively, at the beginning of the experiment. The highest populations were in CSS and FGM soil planted to 'Crimson Sweet,' with populations in CSS soil greater than all others at the second and third planting. The population of F. oxysporum in CSS soil also was greater than in LFC soil throughout the experiment. When planted to 'Florida Giant,' the population of F. oxysporum in CSS soil was not different than that in FGM soil at any planting after initial infestation. There was virtually no detectable population of indigenous F. oxysporum in CSMW soil throughout most of the experiment due to the microwave treatment this soil received. The low levels of F. oxvsporum observed in CSMW soil at plantings four and five represent limited recolonization.

Fusarium wilt levels in soils planted to 'Florida Giant' were high throughout all plantings in conducive soils and tended to increase with the first few successive plantings








79

in the monoculture soils (Figure 4-3). Suppression of Fusarium wilt in CSS soil, which was evident at the first planting (significantly lower wilt than all other soils), was no longer present by the third planting or thereafter. No differences in level of wilt between OSS and the other soils were observed after the third planting and all soils showed disease levels of 70% and greater. Although in this test the disease level in FGM soil did not continue to increase through the fourth and fifth planting of 'Florida Giant,' a repeat of this experiment as well as a similar additional experiment showed increasing wilt through the fourth or fifth planting (Appendix A; Figure A-i and Appendix B; Figures B-i to B-3). In contrast, Fusarium wilt remained low (20-27% wilt) throughout the five successive plantings of 'Crimson Sweet' in CSS soil and also did not increase in FGM soil, although initial levels were higher (34-42%) than those in OSS soil (Figure 4-3). In LEC and CSMW soils, wilt levels were high in the first planting, but did not change dramatically with successive plantings (46-66% and 65-83% for LFC and CSMW soils, respectively). Differences in wilt suppression among the soils were evident through each successive planting, with CSS soil maintaining lower disease levels than those in LEO or CSMW soils. A repeat of this experiment showed similar results regarding E. oxvsporum f. sp. niveum, indigenous F. oxysporum, and wilt levels, with the exceptions already noted and that of showing declining wilt levels in LFC soil successively planted to 'Crimson Sweet' (Appendix B; Figures B-2 and B-5).


Root Colonization by Fusarium oxvs~orum


Root surface colonization measured after four successive plantings of 'Crimson Sweet' and 'Florida Giant' revealed differences among the soils (Table 4-1). Colonization of 'Crimson Sweet' by f. oxysporum f. sp. niveum was lowest in the two monoculture








80

soils, but was not different between CSS and FGM soil. Root colonization by the pathogen was greater in LFC and CSMW soils than in both monoculture soils, with CSMW soil demonstrating the highest level of colonization. Colonization by indigenous F. oxysporum was several times that by F. oxysporum f. sp. niveum in both CSS and FGM soils, both of which had greater levels of colonization by F. oxysporum than those in LFC and CSMW soils. Colonization by F. oxvsporum was lowest in CSMW soil, but did show substantial root colonization due to recolonization of the soil over time after microwave treatment. The ratio of colonization by the pathogen to colonization by indigenous F. oxysporum also was similar in CSS and FGM soils and was less than in the two conducive soils. Root surface colonization by the pathogen was correlated (r=.71) to soil population levels of the pathogen at the time of root sampling. Colonization by E. oxysporum also was correlated (r=.69) with its respective soil populations as was observed previously (Chapter 3). Colonization by the pathogen was negatively correlated (r=-.79) with colonization by E. oxysporum. Wilt was lower in CSS than in LFC and CSMW soils with cultivar 'Crimson Sweet.' Wilt levels also increased and were correlated (r=.68) with soil populations of the pathogen.

With planting to cultivar 'Florida Giant,' slightly higher levels of colonization by F. oxysporum f. sp. niveum were observed for all soils, and differences between the monoculture soils (CSS and FGM) and the conducive soils were similar to those observed with cultivar 'Crimson Sweet' (Table 4-1). Colonization by indigenous F. oxvsoorum was slightly lower in FGM, LFC, and CSMW soil than with cultivar 'Crimson Sweet,' but followed the same pattern of differentiation as observed previously, as did the colonization ratio of F. oxvsporum f. sp. niveum/F. oxvsoorum. However, wilt was high in all soils with no significant differences among them. Wilt, then, was not directly related to the level of








81

colonization by the pathogen or indigenous F. oxysporum. Soil populations of the pathogen at the time of sampling were lower in the monoculture soils than in conducive soils, and populations of F. oxvsporum were highest in monoculture soils planted to 'Crimson Sweet.' Significant differences were also observed when comparisons were made between cultivars (ANOVA, Duncan's, P<0.05) with colonization by F. oxvsporum f. sp. niveum higher in FGM and LFC soil when successively planted to 'Florida Giant' versus 'Crimson Sweet.' Similar results were observed in a repeat of this experiment (Appendix A; Table A-9).

Internal root colonization by f. oxvsporum f. sp. niveum, measured using surfacedisinfested roots, was not different between CSS and FGM soil, although both monoculture soils had lower colonization by the pathogen than that in LFC soil (Table 42). Colonization by indigenous F. oxysporum was similar between all three soils. The ratio of colonization by the pathogen/F. oxvsporum was similar in the monoculture soils, but lower than in LFC soil. Wilt levels were again significantly different among the three soils; wilt was low in CSS soil, moderate in FGM soil, and severe in LFC soil.


Soil Microorganism Populations


Successive plantings of cultivar 'Crimson Sweet' in CSS, FGM, and CSMW soil resulted in increases in populations of bacteria, actinomycetes, fluorescent pseudomonads, and other pseudomonads compared to unplanted soil (Figure 4-4). Increases in populations were observed in LFC soil for actinomycetes, fluorescent pseudomonads, and other pseudomonads, but not for overall bacteria. In FGM soil, population increases following planting to 'Crimson Sweet' also were greater than after planting to 'Florida Giant' for all microorganism groups except fungi. Populations of








82

actinomycetes and fluorescent pseudomonads also were greater in 055 and CSMW soils when planted to 'Crimson Sweet' than to 'Florida Giant.' Overall, in all soils, planting watermelon ('Florida Giant' or 'Crimson Sweet') caused increases in the general prokaryotic microorganism populations as measured in this study. Planting watermelon had no effect on overall fungal populations in any soil, except LFC, in which overall fungal populations decreased when successively planted to 'Crimson Sweet' or 'Florida Giant.' Fungal populations also were lower when planted to 'Crimson Sweet' than to 'Florida Giant' in this soil. LFC soil had the highest overall fungal populations and lowest bacterial populations initially.


Discussion


In this study, cultivation of watermelon, as well as the particular cultivar planted, had significant effects on the populations of F. oLxvscorum pathogenic and not pathogenic to watermelon. When planted to the susceptible cultivar 'Florida Giant,' pathogen populations tended to increase with successive planting, while populations of indigenous F. oxysporum did not change significantly. When planted to 'Crimson Sweet,' pathogen populations in all soils did not differ substantially from those observed in unplanted soils (Chapter 3), regardless of the number of plantings. Populations of indigenous F. oxysgorum, however, tended to increase overall with successive planting to 'Crimson Sweet' in the field soils. Thus, 'Crimson Sweet' appeared to selectively favor the growth of nonpathogens over pathogens, whereas 'Florida Giant' tended to promote pathogen development over nonpathogens. Several studies have demonstrated that susceptible crops increase pathogen populations, whereas nonhost, or resistant hosts, even when






























Figure 4-1. Population dynamics of Fusarium oxvsporum f. sp. niveum (as represented by an orange-colored mutant pathogen) in four soils with successive plantings of two different watermelon cultivars (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil; LFC=fallow, conducive soil; CSMW=suppressive soil rendered conducive by microwave treatment). A) Watermelon cultivar 'Florida Giant' (susceptible to Fusarium wilt); B) watermelon cultivar 'Crimson Sweet' (moderately resistant to Fusarium wilt and inducer of soil suppressiveness). Population estimates were made at the time of planting for each planting number. Values within each planting for each cultivar topped by the same letter are not significantly different (~<0.05) according to Duncan's multiple range test.








84







c A Florida Giant
C
f
U

2000
g
S 1600
o

1200


800o CSMW
LFC
400
FGM

0 CSS
0 1 2 3 4 5
Planting number


B Crimson Sweet
f
U

2000
g
S 1600
0
1200


800 CSMW
LFC
400
FGM
0 cSS
0 1 2 3 4 5
Planting number































Figure 4-2. Population dynamics of indigenous Fusarium oxvsporum in four soils with successive plantings of two different watermelon cultivars (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil; LFC=fallow, conducive soil; CSMW=suppressive soil rendered conducive by microwave treatment). A) Watermelon cultivar 'Florida Giant' (susceptible to Fusarium wilt); B) watermelon cultivar 'Crimson Sweet' (moderately resistant to Fusarium wilt and inducer of soil suppressiveness). Population estimates were made at the time of planting for each planting number. Values within each planting for each cultivar topped by the same letter are not significantly different (p








86





A Florida Giant
C
f
U

4000
g
3500
a
o 3000
I

I 2500
2000
1500 CSS
1000 FGM
500 LFC
0 CSMW
0 1 2 3 4 5
Planting number


B Crimson Swe
C
f
U
/
4000
g
3500
s
o 3000

I 2500
2000
1500 CSS
1000 FGM
500 LFC
0 CSMW
0 1 2 3 4 5
Planting number






























Figure 4-3. Fusarium wilt in four soils with successive plantings of two different watermelon cultivars (CSS=suppressive, monoculture soil; FGM=nonsuppressive, monoculture soil; LFC=fallow, conducive soil; CSMW=suppressive soil rendered conducive by microwave treatment). A) Watermelon cultivar 'Florida Giant' (susceptible to Fusariumn wilt); B) watermelon cultivar 'Crimson Sweet' (moderately resistant to Fusarium wilt and inducer of soil suppressiveness). Values within each planting topped by the same letter are not significantly different (~<0.05) according to Duncan's multiple range test.















Florida Gi
P
e
r
c
e
n 100
t
w 80
i

t 60


40 CSMW

20 LFC
20
FGM

0 ,, / ,, css
1 2 3 4 5
Planting number


B Crimson Sweet
P
e
r
c
e n 100
t
w 80
i
I
t
60


40 CSMW
LFC
20
20 FGM

0 -I CSS
1 2 3 4 5
Planting number