UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanisms of biological control of crown and root rot in tomato by a nonpathogenic Fusarium oxysporum… Sircom, Katharine M. 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_sircom_katharine.pdf [ 7.53MB ]
Metadata
JSON: 831-1.0086176.json
JSON-LD: 831-1.0086176-ld.json
RDF/XML (Pretty): 831-1.0086176-rdf.xml
RDF/JSON: 831-1.0086176-rdf.json
Turtle: 831-1.0086176-turtle.txt
N-Triples: 831-1.0086176-rdf-ntriples.txt
Original Record: 831-1.0086176-source.json
Full Text
831-1.0086176-fulltext.txt
Citation
831-1.0086176.ris

Full Text

MECHANISMS OF BIOLOGICAL CONTROL OF CROWN AND ROOT ROT IN TOMATO BY A NONPATHOGENIC FUSARIUM OXYSPORUM STRAIN  BY KATHARINE MARY SIRCOM B.A., The University of King's College, 1987  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in THE FACULTY OF GRADUATE STUDIES (Department of Plant Science)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA December 1992 © Katharine Mary Sircom, 1992  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of ^PI ct^Sc-^ The University of British Columbia Vancouver, Canada Date Ste- . Co w'" Ict q  DE-6 (2/88)  3  ABSTRACT  The biological control of crown and root rot in tomato was studied in a sterile system in which tomato seedlings were grown on water agar in petri plates or on filter wicks in test tubes containing fertilizer solution. The biological control agent, nit Bl, was a nonpathogenic strain of Fusarium oxysporum which had been mutated to a nitrate non-utilizing form in order to distinguish it from the pathogen, Fusarium oxysporum f. sp.  radicis-lycopersici (FORL). Four different approaches were used to deduce possible mechanisms of control. One approach was to study the effect of inoculum timing and inoculum density on biological control. Biological control was demonstrated when nit B1 was inoculated to the roots as much as 21 days before FORL, and persisted for at least 23 days after the FORL inoculation. When nit B1 was inoculated to seedling roots at least 3 days before FORL, there was good disease control even when the initial inoculum density of nit B1 was 60 times lower than that of FORL. As the lag between inoculations decreased, higher densities of nit B1 were needed to bring about control, and when the two fungi were inoculated simultaneously the inoculum density of nit B1 had to be at least 10 times that of FORL for there to be any reduction of disease symptoms. The second approach was to observe the colonization of seedling roots by the two strains. Both nit B1 and FORL colonized the outer layers of the root. However, when nit B1 was inoculated to the root 4 days before FORL, the rate of increase of nit B1  ii  was greater than the rate of increase of FORL. The third approach was to test possible elicitors of a defence reaction in tomato for their biological control ability. Sterile filtrates from cultures of nit B1 grown in nutrient broth, sterile exudates from nit B1-infested germinating seeds and seedling roots, heat-killed nit B1 spores, and the cell wall fraction from nit B1 all failed to protect seedlings against crown and root rot induced by FORL. The fourth approach was to test nutrient competition by adding an excess of nutrients that might otherwise be limiting. Biological control by nit B1 was not affected when excess glucose or iron were added to the growth medium. In a related experiment, FORL caused severe disease symptoms (in the absence of nit B1) even when iron availability was artificially decreased by adding a strong iron chelator to the growth medium. The conclusion from all these experiments was that nit B1 may elicit a defense response in tomato roots, possibly dependent on the prior colonization of the roots by this strain, which makes the roots resistant to subsequent infection by FORL.  iii  TABLE OF CONTENTS Page Abstract^  ii  List of Tables^  vi  List of Figures^  viii  Acknowledgement Introduction^  1  Literature Review^  2  Chapter 1 Characterization of Biological Control of Fusarium oxysporum f.sp. radicis-lycopersici by Nonpathogenic Fusarium oxysporum Isolate Nit B1 Introduction^ Materials and Methods^ Results^ Discussion^  22 25 38 51  Chapter 2 Colonization of Tomato Seedling Roots by Nonpathogenic Fusarium strain Nit B1 and Fusarium oxysporum f.sp. radicis-lycopersici Introduction^ Methods^ Results^ Discussion^  60 63 66 72  Chapter 3 Testing Induced Resistance as a Possible Mechanism of Biological Control Introduction^ Methods^ Results^ Discussion^  77 81 85 90  iv  Table of Contents (con't)^  Page  Chapter 4 Testing Nutrient Competition as a Possible Mechanism of Biological Control Introduction^ Methods^ Results^ Discussion^  93 96 99 106  General Discussion^  111  Bibliography^  117  Appendix A - ANOVAs^  126  Appendix B - Media Recipes^  143  Appendix C - Physiological Phenotypes of Nit Mutant Strains^  148  v  LIST OF TABLES Chapter 1  Page  Table 1.1. Sources of pathogens and nonpathogenic  strains  27  Table 1.2. Mean disease severity ratings of tomato  seedlings inoculated with nitrate non-utilizing (nit) mutant or wild-type Fusarium strain B1 and then with Fusarium oxysporum f. sp. radicis -lycopersici (FORL)  41  Chapter 2 Table 2.1. Colony forming units of Fusarium strain  nit B1 and Fusarium oxysporum f. sp. radicislycopersici (FORL) recovered from previouslyinoculated tomato seedling roots  67  Chapter 3 Table 3.1. Mean disease severity ratings of plants  treated with possible elicitors from broth cultures of Fusarium strain nit B1 and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL)  86  Table 3.2. Mean disease severity ratings of plants  treated with possible elicitors from nit B1infested germlings and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL)  87  Table 3.3. Mean disease severity ratings of plants  treated with possible elicitors from nit B1infested roots and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL)  88  Table 3.4. Mean disease severity ratings of plants  treated with the heat-released cell wall fraction of Fusarium strain nit B1 and then inoculated with Fusarium oxysporum f. sp. radicislycopersici (FORL)  vi  89  List of tables (con't) ^  Page  Chapter 4 Table 4.1. Biological control of Fusarium crown  and root rot by nonpathogenic Fusarium strain nit B1 in the presence of excess glucose and nitrogen ^100  Table 4.2. Biological control of Fusarium crown  and root rot by nonpathogenic Fusarium strain nit B1 in the presence of excess iron ^  101  Table 4.3. Mean root weights (mg) of tomato  seedlings grown in iron-rich media, inoculated with nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici ^ (FORL)  103  Table 4.4. Colony forming units of nonpathogenic  Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici washed from the roots of tomato seedlings grown in iron-rich media ^  103  Table 4.5. Mean disease severity ratings of tomato  seedlings treated with the iron chelator EDDHA and then inoculated with Fusarium oxysporum f. ^ sp. radicis-lycopersici (FORL)  vii  105  LIST OF FIGURES Chapter 1^  Page  Figure 1.1. Growth of nitrate non-utilizing (nit)  mutants of biological control Fusarium strains and the wild-type form of B1 on Fusarium minimal medium (FMM)^  Figure 1.2. Growth of tomato seedlings in  different fertilizer solutions^  28 34  Figure 1.3. Disease severity rating scale for  tomato seedlings inoculated with Fusarium ^ oxysporum f. sp. radicis-lycopersici  35  Figure 1.4. Water-soaked stem and leaf lesions on  tomato seedlings infested with biological control Fusarium strain nit Bl. ^  39  Figure 1.5. Effect of Fusarium oxysporum f. sp.  radicis-lycopersici (FORL) inoculum density on disease severity ratings of unprotected tomato seedlings  42  Figure 1.6. Mean disease severity ratings, on a  scale of 0 (healthy) to 5 (dead) of tomato seedlings inoculated with nonpathogenic Fusarium strain nit B1 and then with Fusarium oxysporum f. sp. radicis-lycopersici (FORL) after increasing time intervals ^  43  Figure 1.7. Effect of nonpathogenic Fusarium  strain nit B1 inoculum density on severity of crown and root rot in tomato seedlings induced^ by Fusarium oxysporum f. sp. radicis-lycopersici 45  Figure 1.8. Effect of increasing time intervals  (from 0-4 days) between inoculation with the nonpathogenic Fusarium strain nit B1 and inoculation with Fusarium oxysporum f. sp. radicislycopersici on severity of crown and root rot in tomato seedlings ^  46  Figure 1.9. Effect of different inoculum density  ratios of nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) on severity of crown and root rot symptoms in tomato seedlings^  viii  48  List of Figures (con't)^ Figure 1.10. Effect of total inoculum density [nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL)] on severity of crown and root rot symptoms in tomato seedlings. ^  Page  50  Chapter 2 Figure 2.1. Colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) recovered from unrinsed tomato roots over 10 days ^  69  Figure 2.2. Colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici washed from the surface (A) or recovered from rinsed, ground tomato roots (B) over 20 days ^  70  ix  ACKNOWLEDGEMENT  I would like to thank my research supervisor, Dr. R. J. Copeman, for his kind patience and support of this thesis. Thanks also go to the members of my supervisory committee: Drs. S. M. Berch, H. S. Pepin and B. E. Ellis, for their comments and criticisms. I acknowledge summer students Tanis Douglas and Alex Bell for their help in the lab. And I thank all the students in the Plant Science Department for their friendship and kindness over the past three years.  x  INTRODUCTION  Crown and root rot in tomato is a fungal disease caused by Fusarium oxysporum f. sp. radicis-lycopersici (FORL). The disease is a problem in temperate tomato growing regions around the world, and, locally, has caused considerable losses in the Fraser Valley greenhouse tomato industry. Fusarium oxysporum diseases are particularly suited to biological control and tomato crown and root rot is no exception: economically significant disease reduction by nonpathogenic Fusarium oxysporum strains has been demonstrated in France, Ontario and British Columbia. There are still problems in using this form of control commercially. In particular, we do not yet know the mechanisms by which the nonpathogenic strains bring about the control so we do not know what variables to change to enhance it. The purpose of this research was to develop a system for studying the interaction of nonpathogenic F. oxysporum strains and FORL in vitro and to suggest some mechanisms for the observed disease control by the nonpathogenic strains.  1  LITERATURE REVIEW  The Genus Fusarium  Members of the genus Fusarium are among the most common and widely distributed fungi on earth. They are mainly soil inhabitants, but they can be easily isolated from air, water and diverse organic substrates. Some reasons for their ubiquity are their efficient mechanisms for dispersal and survival and a capacity to change their morphology and physiology in response to changes in their environment (Booth, 1971; Burgess, 1981). The defining characteristic of the genus Fusarium is the presence of fusiform (spindle-shaped) macroconidia which have a foot cell bearing some sort of heel (Booth, 1971). Species within the genus are distinguished by the exact shape of their macroconidia and by the manner in which their macroconidia, microconidia, and chlamydospores are produced (Tousson and Nelson, 1976). The species which is the most economically important is  Fusarium oxysporum, for about half its members are root-infecting pathogens (Booth, 1971). Most of these pathogens cause vascular wilts: the fungus invades the root and proceeds to the stele, then spreads systemically in the vascular stream, interfering with the flow of water and nutrients and eventually causing a wilt in the upper parts of the plant (Beckman and Talboys, 1981). A few Fusarium oxysporum pathogens cause root rots: they are 2  restricted to the roots, crown and lower stem where they cause extensive damage to the cortical tissue and the parenchyma tissue of the stele, but seldom spread systemically in xylem vessels (Beckman and Talboys, 1981).  Fusarium oxysporum pathogens are divided into formae speciales on the basis of host specificity. At least 76 of these have been described. Some of the more serious ones are F.  oxysporum f. sp. melonis which causes vascular wilt of muskmelon and canteloupe, F. oxysporum f. sp. dianthi which causes vascular wilt of carnations, and F. oxysporum f. sp. cucumerinus which causes vascular wilt in cucumber. F. oxysporum f. sp. lycopersici causes a vascular wilt in tomatoes which has in the past been a serious problem: it is controlled now by the use of resistant cultivars. Other isolates of Fusarium oxysporum have no known hosts. These nonpathogenic strains are commonly isolated from the rhizosphere of healthy plants (Taylor and Parkinson, 1961; Gordon et al. 1989). Little is known about the ecological role of these strains.  Fusarium Crown and Root Rot in Tomato  Crown and root rot in tomato was first reported in Japan in 1969 (Sato and Araki, 1974). In the 1970's it became a problem in North America (Leary and Endo, 1971; Sonoda, 1976; Nutter et al., 1978) and in the 1980's spread to Europe (Hartman and Fletcher, 1991; Lemanceau and Alabouvette, 1991). The causal agent was 3  originally thought to be a new race of the tomato vascular wilt pathogen, F. oxysporum f. sp. lycopersici (FOL), but its symptomology and host range were so different that in 1978 it was deemed a new forma specialis, named F. oxysporum f. sp. radicislycopersici (FORL) (Jarvis and Shoemaker, 1978). Today, Fusarium crown and root rot is probably the most serious disease in greenhouse tomato crops wherever they are grown. The first symptom of the disease is a yellowing, necrosis and wilt of the lower leaves that generally appears just as the first fruits are ready for picking. In some plants this symptom progresses slowly up the plant and there is a gradual wilt as the infected parts collapse; in others the first symptoms are followed by a sudden wilt of the whole plant and, soon after, death. Young infected plants damp off (Jarvis, 1988). When diseased plants are pulled up, a dark brown cortical rot is visible at the crown and there may be dark lesions on the secondary roots. Often the tap root is completely rotted away. Although there is sometimes a reddish-brown vascular discolouration extending up the lower part of the stem, the fungus itself (in contrast to the vascular wilt fungus) cannot usually be isolated from more than a few millimeters away from the root lesions (Jarvis, 1988). F. oxysporum f. sp. radicis-lycopersici infects most commercially grown tomato cultivars, including those that are resistant to F. oxysporum f. sp. lycopersici (Jarvis and Thorpe, 1976). It is mainly a problem in greenhouses, but it has been 4  found in field-grown crops as well (Leary and Endo, 1971; Sonoda, 1976; Brammall and McKeown, 1989). The fungus can also cause severe root lesions in species of the Leguminosae, Cucurbitaceae and Chenopodiaceae (Rowe, 1980; Menzies et al., 1990).  Infection and Pathogenesis  The infection of tomato seedlings has been studied microscopically (Charest et al., 1984; Brammal and Higgins, 1988). Microconidia of Fusarium oxysporum f. sp. radicis-  lycopersici (FORL) germinate about 12 h after inoculation and hyphae grow over the surface of the root before putting out infection pegs that either penetrate the epidermal cells directly, or else grow through the middle lamella between two adjoining cells. From 24 to 72 h after inoculation the fungus colonizes the epidermal cells and attempts to penetrate the first layer of cortical cells (the hypodermis). Cells directly adjacent to fungal hyphae respond to infection: there are noticible thickenings in wall structure, and papillae may form in the space between the cell wall and the cytoplasm. In a resistant reaction, these cell changes (possibly in combination with other defense mechanisms) are successful in limiting the fungus to the epidermis and the hypodermis. In a susceptible plant, however, the fungus can actually penetrate papillae (possibly because the formation of these defense structures has been delayed, so they are not fully mature at the time of fungal invasion). Once 5  through the hypodermis, the fungus rapidly colonizes the inner cortex, causing extensive degradation of this tissue. Between 96 and 120 h after inoculation the fungus penetrates the endodermal cells and reaches the stele, where it invades parenchymous cells, primary phloem cells and protoxylem vessels. These cells react in a number of ways, but once colonized they eventually disintegrate. Colonized xylem vessels accumulate a fibrogranular coating material interspersed with small bubble-like structures, and tyloses may form in cells above those that are colonized; these seem to limit the upward spread of the pathogen.  Comparison of FORL and FOL  Although the outward symptoms of Fusarium crown and root rot are different from the outward symptoms of Fusarium wilt, it is important to realize that root colonization by the two pathogens is somewhat similar (Beckman and Talboys, 1981; Bishop and Cooper, 1983). F. oxysporum f. sp. lycopersici (FOL) germinates in about the same time and again grows on the outside of the root. Invasion is only through the middle lamella between two cells, and not directly through the tangential wall of the epidermis. Colonization of the cortex is much slower and less extensive than it is with FORL, possibly because of an earlier host reaction, leading to more effective deposition of wall barriers. Unlike FORL, FOL never penetrates the papillae that are formed in plant cells adjacent to the invading hyphae - it only 6  invades cells that do not have visible wall modifications. The cortex is eventually traversed, mainly by intercellular growth. Interestingly, this limited invasion of the cortex occurs in both resistant and susceptible hosts, and has also been observed with nonpathogenic F. oxysporum species (Bishop and Cooper, 1983). Rather than occurring at the hypodermis, resistance to the wilt pathogen seems to be determined at the endodermis, which accumulates a large amount of electron-opaque material and limits (but does not completely prevent) invasion of the stele. Vessel elements respond to invasion by FOL in much the same way as they do to invasion by FORL: there is an accumulation of coating material and a formation of tyloses in higher vessels. These limit the spread of FOL in the incompatible reaction but fail to do so in the compatible reaction. From a comparison of infection by these two pathogens it could be argued that FORL behaves exactly like FOL in a susceptible plant. The only difference, on a structural level, is that FORL completely overcomes resistance mechanisms in the cortex, so that it can extensively colonize this tissue, while FOL is limited in the cortex and only multiplies rapidly when it reaches the stele. Other comparisons of FORL and FOL have been made: the two formae speciales have different protein compositions (Belhadj et al., 1987) and different quantities of lectin-binding sugars on their mitochondria (Boyer and Charest, 1989); they also have differences in their fatty acid composition and mycostatin 7  tolerance (Madhosingh and Starratt, 1987)  Control of Fusarium Crown and Root Rot  Conidia of FORL are airborne, so early attempts to control the fungus by greenhouse sterilization were unsuccessful, as it readily recolonized the disinfested growth medium (Rowe et al., 1977). Chemical treatments have been recommended in some places (Mihuta-Grimm et al., 1990), but these have not been used widely in Canada (Jarvis and Thorpe, 1976). There are some resistant cultivars, but these are not used widely in commercial production (Copeman, personal comm.). At present most growers in British Columbia are controlling the disease by grafting onto resistant root stock (Copeman, personal comm.; Thorpe and Jarvis, 1981). This is labour-intensive and, needless to say, there is considerable interest in finding alternative controls.  Biological Control of Fusarium Diseases  Fusarium oxysporum pathogens are particularily amenable to biological control. In the first place, the losses they cause are generally due to a reduction in yield as the plants become stressed and die, rather than a complete loss as is the case with pathogens that attack the fruit and render it unmarketable. A form of control which may not completely eradicate the pathogen, but rather delays or reduces its impact is therefore acceptable. 8  Secondly, the soil is naturally home to a large number of saprophytes and predators and the rhizosphere, in particular, supports an enormous number of nonpathogenic bacteria and fungi, some of which enter into a symbiotic relationship with the plant (Foster et al., 1983). Until the root-infecting fungus actually enters the root and occupies its own particular niche it is subject to all the influences of these other microorganisms. Many of them may have a much higher competitive saprophytic ability than the pathogen (Marois et al., 1981; Jarvis, 1989). A great deal of research has been done recently on the biological control of soil-borne pathogens and, not surprisingly, much of it has focussed on Fusarium oxysporum. Many different biological control agents have been shown to reduce losses caused by this species. They can be organized into four groups. The first group is plants: this is called allelopathy. For example, lettuce and dandelion can alleviate the effects of FORL when interspaced with tomatoes (Jarvis, 1988). The second group is fungi that are unrelated to F. oxysporum. An example is Trichoderma harzianum which has reduced disease incidence (Marois et al., 1981) and increased yields (Sivan et al., 1987) of tomato grown in FORL-infested fields. The presence of a vesicular-arbuscular mycorrhizal (VAM) fungus, Glomus intraradices on tomato roots may also limit the development of Fusarium crown and root rot (Caron et al., 1986). The third and fourth groups of biological control agents are associated with Fusarium wilt-suppressive soils. In suppressive 9  soils, a particular disease is very restricted or does not occur at all, even if the pathogen is present and conditions are favourable for its growth (Alabouvette et al., 1979). The cause of the suppressiveness is biological: for reasons which are still not clear these soils favour a community of microorganisms which, separately or in association, interfere with the ability of the pathogen to cause disease. The microorganisms that are involved in the suppressiveness vary from place to place. Research on Fusarium wilt-suppressive soils in California has focussed on fluorescent Pseudomonas species, the third group of biological control agents. Various strains have been isolated from suppressive soils; when transferred to disease-conducive soils they can make them suppressive (Schroth and Hancock, 1982). Over the past ten years, these pseudomonads have given good biological control of Fusarium wilt diseases of flax, cucumber, radish, and carnation both in pot trials and in the field (Kloepper et al., 1980; Scher and Baker, 1980; Yuen et al., 1985). Members of the genus Fusarium make up the fourth group of organisms showing biological control activity against Fusarium wilt pathogens. The Fusarium wilt-suppressive soils that are found in the Chateaurenard region of the Rhone Valley in France support high numbers of nonpathogenic Fusarium oxysporum strains and these are thought to be mainly responsible for the suppressiveness (Louvet et al., 1981). Like fluorescent pseudomonads, nonpathogenic Fusarium oxysporum strains control a 10  variety of Fusarium wilts when transferred into conducive soils (Paulitz et al., 1987; Lemanceau and Alabouvette, 1991). Nonpathogenic fusaria with similar biological control abilities have also been isolated from suppressive soils in California (Paulitz et al., 1987). Other researchers have demonstrated biological control by avirulent strains of the pathogen in question (Mas et al., 1981), or else by formae speciales that are pathogenic on a different host (Kroon et al., 1991). Some biological control agents are effective against a variety of pathogens. For example, some fluorescent pseudomonads that control Fusarium wilts also control take-all of wheat caused by Gaeumannomyces graminis (Kloepper et al., 1980). Also, biological control agents can work in concert: recent papers document improved biological control against Fusarium wilt of carnation and cucumber, and against Fusarium crown and root rot of tomato when a fluorescent pseudomonad from California is combined with a nonpathogenic strain of F. oxysporum from France (Park et al., 1988; Lemanceau et al., 1992).  Biological Control Against FORL in Canada  A selection of nonpathogenic F. oxysporum and F. solani strains isolated from disease-free tomato plants in Ontario have shown promise as biological control agents against Fusarium oxysporum f. sp. radicis-lycopersici. Tomato plants inoculated with combinations of these strains had significantly higher 11  yields than non-protected plants both in a hydroponic system in Ontario (Louter and Edgington, 1990) and in sawdust culture here in British Columbia (Copeman et al., 1990). The increase in mean yields was high (27% in Ontario and 35% in British Columbia) but of special importance is the fact that in both cases there was a selective increase in the yield of extra-large fruit. Because these fruit have a higher market value, the economic benefit of biological control is potentially even greater than the increase in yields suggests (Copeman, personal comm.).  Mechanisms of Biological Control  Mechanisms of biological control are poorly understood, but there is evidence in the literature for at least four general types. One is antibiosis, here used for examples in which the biological control agent has a direct deletarious effect on the pathogen. Antibiosis can be demonstrated in vitro. For example, Trichodema harzianum can destroy fungal hyphae in dual culture assays (Cherif and Benhamou, 1990) and some bacteria produce antibiotics which are toxic to Fusarium wilt pathogens (Sivamani and Gnanamanickan, 1988). Phenolic compounds found in some allelopathic plants inhibit the growth of FORL in culture (Kasenberg and Traquair, 1988), so this too, in a sense, could be called antibiosis. The importance of any of these direct antagonistic effects has yet to be proven in vivo. There is no 12  evidence that direct antibiosis is involved in the biological control of Fusarium wilt diseases by nonpathogenic Fusarium oxysporum strains. The second mechanism is that in which the biological control agent elicits a systemic or partially systemic resistance response in the host. This is sometimes called "cross-protection" because in many of these cases the biocontrol agent is an avirulent strain of the pathogen or an alien pathogenic fungus, so the protection can be interchangeable: what was once the "pathogen" becomes the inducer of resistance when a different host is used. It is perhaps clearer to use the term "induced resistance" (Matta, 1989). Induced resistance against F. o. lycopersici has been achieved by inoculating tomato plants with F. o. dianthi (Wymore and Baker, 1982; Kroon et al., 1991) and by inoculating Verticillium-resistant tomato plants with Verticillium dahliae (Jorge et al., 1992). Similarily, resistance against Fusarium wilt of muskmelon and watermelon has been induced by avirulent races of F. o. melonis and F. o. niveum respectively (Mas et al., 1981; Biles and Martyn, 1989), and resistance against Fusarium wilt of sweet potato has been induced by nonpathogenic Fusarium oxysporum strains (Ogawa and Komada, 1986). Induced systemic resistance is proven experimentally in these systems by inoculating the roots with the inducing organism and inoculating the stem with the pathogen (Ogawa and Komada, 1986) or by splitting the roots and inoculating one half with the inducer and 13  the other half with the pathogen (Kroon et al., 1991). The extent of biological control achieved in these systems varies. In some, the resistance is transitory and differs little from the nonspecific, short-term resistance that is induced in plants subjected to any one of a number of stresses, including abiotic stresses (Mas et al., 1981; Wymore and Baker, 1982; Kroon et al., 1992). In others, the protection lasts longer (Jorge et al., 1992). The degree of protection may in part depend on the inducing organism, but the inoculum densities of inducer and challenger and the timing of inoculum may all affect the outcome (Jorge et al., 1992). The third mechanism of biological control is competition for nutrients. Exogenous sources of carbon and nitrogen are necessary for germination of Fusarium conidia when spore densities are high (Griffin, 1981). One theory for the suppressivness of soils in the Chateaurenard is that the microflora (including the nonpathogenic fusaria) in these soils act as a sink for the carbon and nitrogen in root exudates, making them unavailable for pathogen chlamydospore germination. Some experiments have given support to this theory: for example, adding glucose to a suppressive soil increased germination of F. o. melonis chlamydospores and nullified the suppression (Louvet et al., 1981). In a different system, adding glucose and asparagine or excess seedling exdates nullified control of F. o. vasinfectum and F. o. melonis by Trichoderma harzianum (Sivan and Chet, 1989). This same Trichoderma harzianum had earlier been an 14  effective biocontrol agent against FORL (Sivan et al., 1987). Although carbon competition may be important in soils in which all the soil microflora contribute to the suppression, there is less evidence for it in intraspecific biological control (Alabouvette, 1990). Paulitz (1990) has pointed out that competition for carbon between Fusarium oxysporum strains is sometimes suggested as a mechanism of control when no other mechanism can be proven. In suppressive soils in which fluorescent pseudomonads are the main biological control agent competition for iron may be more important than competition for carbon. Iron is an important element in germ-tube elongation and infection by Fusarium pathogens (Simeoni et al., 1987). Fluorescent pseudomonads produce a siderophore with a higher affinity for iron than the siderophores produced by Fusarium species; thus the bacteria render iron unavailable to the pathogens (Scher and Baker, 1982). This mechanism is supported by a number of different kinds of experiment. Adding excess iron to the growth medium, in the form of FeC1 3 or the low-affinity chelator FeEDTA nullified the control of F. o. cucumeris and F. o. lini by Pseudomonas putida (Kloepper et al., 1980; Scher and Baker, 1982; Sneh et al., 1984). Likewise, lowering the pH, which increases the amount of available iron in the soil reduced control of F. o. cucumerinum by the same biocontrol agent (Park et al., 1988). On the other hand, adding the high-affinity iron chelator FeEDDHA enhanced control of these same pathogens and, when applied on its own, 15  mimicked the effect of the bacterium (Scher and Baker, 1982). Control of wilt was also demonstrated when a siderophore was extracted from a fluorescent pseudomonad and added to wiltconducive soil (Kloepper et al., 1980). A direct correlation between siderophore production by various pseudomonads and their control of F. o. cucmerinum (as measured by inhibition of chlamydospore germination) has been noted (Sneh et al., 1984). More recently, experimenters have used Tn5 transposon mutagenesis to obtain Pseudomonas mutants defective in the production of siderophores. When the mutants are compared to wild-type strains in biological control assays the wild types give better control of the disease being studied (Baker, 1990; Lemanceau et al., 1992). Adding carbon to soils in which iron competition is a possible mechanism does not nullify control; indeed it has been shown to enhance control of F. oxysporum formae speciales (Elad and Baker, 1985). Adding carbon in this case may have provided extra energy for production of siderophores by the pseudomonad. A similar explanation was given for the enhanced control of F. o. cucumerinum by a nonpathogenic Fusarium strain in the presence of Pseudomonas putida: it was suggested that the nonpathogenic strain stimulated root exudation, thus providing more carbon for the pseudomonad (Park et al., 1988). This explanation was not substantiated, however, and it obviously conflicts with theories about why nonpathogenic Fusarium strains can provide protection on their own. 16  Competition for iron is not usually suggested in cases in which only nonpathogenic fusaria are involved in the suppressiveness. However, one group of researchers comparing amounts of the Fusarium siderophore fusarine in nonpathogenic and pathogenic strains found that the strains that were the most effective biocontrol agents (including Fo47, an isolate that was later shown to be effective against FORL) also synthesized the most fusarine in vitro (Lemanceau et al., 1986). These researchers suggested that iron competition might at least be a factor in soil suppressiveness by nonpathogenic fusaria. Given the recent reports of interactions between siderophore-producing pseudomonads and nonpathogenic fusaria, iron competition may be shown to be more important in these French soils than was previously thought. The fourth mechanism of biological control is that in which the controlling agent somehow makes infection sites on the host unavailable to the pathogen. This is sometimes called "competition for infection sites" (Matta, 1989; Mandeel and Baker, 1991). The biological control agent could conceivably prevent infection by the pathogen at these sites simply by blocking the way. It is more likely, though, that this agent induces a localized defense response in cells directly adjacent to it, or else that it competes for substrates that may be limited at that microsite (as distinct from general nutrient competition in the rhizosphere). Experimentally, this mechanism of protection is difficult to 17  prove, unlike an induced systemic reaction in which the inducer and challenger can be separated spatially. Schneider (1984) compared inoculum density/root colonization curves (using F. o.  apii and wilt-susceptible celery) in the presence of various nonpathogenic fusaria. He determined the hypothetical maximum number of pathogen colonies per unit of root using a data transformation generally used in studies of enzyme/substrate kinetics. Since this hypothetical value was not greatly affected by the presence of different nonpathogenic strains he suggested that potential colonization sites for the pathogen were limited. However, the ability of F. o. apii to occupy these colonization sites (measured by the inoculation density required to get half the maximum root colonization) changed drastically depending on which nonpathogenic strain was used. Schneider therefore concluded that the nonpathogens "affected the susceptibility of a finite number of infection sites" rather than reducing the absolute number of such sites. He gave no suggestion of how these infection sites were affected, though, and it seems possible that the colonization patterns he found could result from a number of mechanisms, including direct antagonism, or competition for nutrients in the larger growing medium. Mandeel and Baker (1991) looked at root colonization by F. o. cucumerinum in the presence of nonpathogenic Fusarium oxysporum strains. They found indications of infection site competition between one of the nonpathogenic strains and FORL: as root colonization by the nonpathogenic strain increased, colonization by the pathogen 18  decreased, and vice versa. It is sometimes difficult to distinguish between different mechanisms of control, and there are examples in which one biocontrol agent can effect control by more than one mechanism. In one such example, partial evidence for nutrient competition, infection site competition and induced systemic resistance was given for the control of Fusarium wilt of cucumber by nonpathogenic fusaria (Mandeel and Baker, 1991). It is important to remember, also, that the types of experiments that are carried out depend on what the experimenters expect the mechanism of protection to be, and further experiments may show that what were once considered to be different mechanisms are actually the same.  The Future of Biological Control Against FORL  Although biological control against F. oxysporum diseases has been extremely successful under experimental conditions, it has had variable success in the field. A common explanation for this is that when biological control agents are introduced into raw soil they fail to compete successfully with the existing microflora to maintain the high population densities that are needed for control of the pathogen (Alabouvette, 1991; Weller, 1988). To avoid this competition, the soil must be disinfested to an unusually deep level before introducing the biological control agents (Alabouvette, 1991). 19  Establishing biological control agents is not a problem in greenhouse crops, since the greenhouses are normally disinfested at the beginning of each growing season. As well, temperature and moisture in greenhouses can be manipulated to favour growth of the biological control agents before the crop is started. For these reasons it will probably be in greenhouses that the first commercial biological control agents against soil-borne pathogens, including FORL, will be applied. In France, a nonpathogenic strain of F. oxysporum, one that has been effective against a number of Fusarium wilt diseases and also against FORL, is in the process of being registered, and work on formulating commercial inocula has begun (Alabouvette, 1991).  Objectives of this Research  The main purpose of this project was to suggest possible mechanisms for the biological control of crown and root rot by nonpathogenic Fusarium oxysporum strains that was demonstrated in British Columbia greenhouse trials. Knowledge of these mechanisms would allow manipulation of growing conditions to enhance the control, and would provide screening criteria for new biological control agents (Alabouvette, 1991). Rather than taking a reductive approach which would involve biochemical studies on the host-nonpathogen or host-pathogen interaction in vitro, a system was developed that involved demonstrating biological control in a living plant. This approach was taken so that an understanding of 20  the ecological relationships of the two fungi on the plant roots could be gained as well. No matter what their mode of action, biocontrol agents must become well-established in the environment to which they are introduced, or their effect will not likely be permanent. Knowledge of the physical activities of the biocontrol agent and the pathogen in the growth medium and on the plant roots is also necessary for planning how and when to apply the biocontrol agents. The first objective in this research was to determine the effect of inoculum density and inoculum timing on biological control, in order to obtain some hints of the mechanisms that might be responsible for the control. The second objective was to describe the basic colonization patterns of the two fungi on tomato roots. The third objective was to determine whether a nonliving elicitor preparation from the nonpathogenic strain could bring about biological control - a positive result would be evidence for induced resistance by the nonpathogenic strain. The fourth objective was to test whether nutrient competition was a factor in biological control in this system.  21  Chapter 1  Characterization of Biological Control of Fusarium oxysporum f. sp. radicis-lycopersici by Nonpathogenic Fusarium oxysporum Isolate Nit B1  INTRODUCTION  The nonpathogenic Fusarium oxysporum isolate B1, that was used as a biological control agent in this study, was one of three isolates obtained from J. H. Louter at the University of Guelph. These isolates were tested in sawdust culture in a commercial greenhouse here in British Columbia in the summer of 1988. In the greenhouse experiment, all three isolates were mixed together and applied to the seeds. At the end of the experiment, plants inculated with the mixture of nonpathogenic isolates had significantly higher yields than the uninoculated controls (Copeman et al., 1990). The three isolates were tested independently in preliminary experiments to this study. Isolate B1 gave the most consistent biological control so it was chosen for further experiments. A nitrate non-utilizing (nit) mutant of isolate B1 was generated to facilitate anticipated root colonization studies, as suggested by Hadar et al. (1989). Nit mutants lack a functional enzyme in the nitrate assimilation pathway, by which nitrate is reduced to ammonium, the form taken up by the fungus. These 22  mutants, unable to use nitrate as a sole nitrogen source, form a thin expansive mycelium when grown on nitrate minimal medium, but form colonies indistinguishable from those of the wild-type strain in the presence of ammonium or organic nitrogen sources. Most nit mutants of Fusarium oxysporum can grow on media that contain chlorate, while wild-type strains cannot. This is presumably because chlorate, a nitrate analogue, is reduced by nitrate reductase to toxic chlorite in the wild-type strains, which is then assimilated by the fungus (Correll et al., 1987). The nit mutants, lacking this functional enzyme (or a cofactor necessary for its activity) are unable to take up the chlorate and are thus unaffected by it. Providing that the nit mutant of B1 does not behave differently from the wild-type strain in biocontrol assays, it can be used in studies of the interaction between pathogen and nonpathogen on the root, using a chloratecontaining medium as a selective medium for it, and nitrate minimal medium as a selective medium for the pathogen. Two experimental systems were used in this study, one in which the seedlings were grown on an agar medium in petri plates and one in which they were grown on paper or glass microfibre wicks in test tubes. The systems were chosen for several reasons: they were simple to set up, they allowed biological control to be quickly assessed, they were sterile, and the plant roots could be easily removed for colonization studies. The first objective of Chapter 1, then, was to test whether the nit mutant of B1 (designated nit 81) was as effective as its 23  wild-type counterpart at protecting plants against FORL in the experimental systems. A second objective of Chapter 1 was to determine some of the parameters of biological control. The effect of FORL inoculum density on disease severity was tested first, partly to find an optimum inoculum density for further experiments and partly to gain some understanding of the disease process. Then the effects of nit B1 and FORL inoculum density and inoculum timing on biological control were tested, to obtain clues about possible mechanisms of biological control.  24  MATERIALS AND METHODS  Fungal Isolates  The isolates of Fusarium oxysporum f. sp. radicislycopersici (FORL) used in this study were collected by Dr. R. J. Copeman from greenhouses in the Fraser Valley in the summer of 1989. Their pathogenicity towards tomatoes was confirmed by a modification of the Sanchez technique (Sanchez et al., 1975). Ten surface-disinfested seeds of a susceptible tomato cultivar were placed in a circle on 1% water agar in a petri plate (25 x 100 mm) and three mycelial plugs (approximately 5 mm diameter) of the fungus to be tested were placed in the center of the plate. The  °  plates were incubated at room temperature (20-25 C) under a benchtop light bank for about 10 days. Fungi that caused distinct dark brown rotting lesions at the crowns of the seedlings were identified as FORL. Three isolates of FORL which were the most virulent in the Sanchez test, and which sporulated consistently on potato dextrose agar (PDA: Difco, Detroit, Michigan) were chosen for this study (Table 1.1). They were single-spored and maintained in  °  the dark at 22 C on acid PDA (PDA with lactic acid to make pH 4.5) throughout the course of the study. Although they were transferred to fresh acid PDA about once a month, they showed no decrease in virulence, as has sometimes been observed with these pathogens (Jarvis, 1988). Spores of the fungi were also stored at 25  4 ° C on a sterile 50:25:25 mixture of soil, peat and vermiculite (Tousson and Nelson, 1976) and occasionally new cultures were started from these. The nonpathogenic strain was obtained from Dr. J. H. Louter at the University of Guelph. He had identified it as Fusarium oxysporum (Table 1.1).  Generation of the Nit Mutant  A chlorate-resistant, nitrate non-utilizing (nit) mutant of B1 was generated by placing mycelial plugs of this isolate on a Fusarium minimal medium (FMM: see recipe in Appendix B) amended with 1.6 g/1 of L-asparagine, as a nitrogen source, and 15 g/1 of potassium chlorate, which is toxic to wild-type Fusarium species (Correll et al., 1987). Within a week, fast-growing, chlorateresistant sectors grew out from the plugs. One sector was transferred to acid PDA, allowed to grow for a week, and then transferred to unamended FMM. It grew on FMM as a clear, thin expansive colony and so was designated a nit mutant (Fig. 1.1). The nit mutant was maintained on acid PDA under the same conditions as the wild-type strain. From time to time it was transferred to plates of FMM or amended-FMM (FMM-C) to check the stability of the mutation.  26  Table 1.1. Sources of pathogens and nonpathogenic strains ^ Source Identity^ Isolate Pathogens: SSa 17-0 pink^Fusarium oxysporum ^R. Copeman, U.B.C. f. sp. radicis-lycopersici DS 89 117 (FORL) ^ ^ II R. Copeman, U.B.C. 16 DS 89 95 ^ ^ II R. Copeman, U.B.0 SSc 33 R 89 204 Nonpathogens: B1^Fusarium oxysporum ^J.H. Louter, Guelph Nit B1^nitrate non-utilizing^generated from B1 mutant of B1  27  Figure 1.1 Growth of nitrate non-utilizing (nit) mutants of biological control Fusarium strains and the wild-type form of B1 on Fusarium minimal medium (FMM). Clockwise from top left: nit FS; nit Bl; nit IPA; wild-type B1. Nit FS and nit IPA were tested in preliminary experiments but not used in the main research.  28  Inoculum Production  Conidial suspensions were prepared by putting a mycelial plug (5 mm diam.) from the fungal colony into 5 ml of sterile distilled water and mixing it on a vortex mixer to suspend the spores. The colonies were usually 2 weeks old when they were used, and they generally sporulated well. Nit B1 and all isolates of FORL produced mostly microconidia. Old cultures of FORL sometimes produced masses of macroconidia as well as microconidia. The inoculum density was measured using a haemocytometer, and was adjusted to the required concentration by ten-fold dilutions into 4.5 ml sterile distilled water blanks. Very little debris or mycelium was seen in the conidial suspensions, so they were never strained through cheesecloth, as is customary.  Petri Plate Assays  Tomato seeds (cultivar Vendor, Stokes Seeds Ltd. St. Catherines, Ont.) were surface-disinfested by soaking in a 1% NaC1O solution (1:5 dilution of household bleach) for 30 minutes, and then rinsing in three changes of sterile distilled water, leaving them for at least 20 minutes in each rinse. In experiments in which the nonpathogen was applied at seeding, 0.5 ml of the nonpathogenic conidial suspension (or sterile distilled water for the controls) was spread onto plates 29  of 1% water agar. Ten surface-disinfested seeds were placed in a circle around the edge of the plate, and the plates were wrapped with Saran-wrap or Parafilm and incubated at 25 ° C in the dark for 3 or 4 days, until the radicle was about 1 cm long. Then the plates were opened and inoculated with the pathogen. Mycelial plugs of FORL were placed in the center of the seeds so that the fungus could grow outwards and infect the crowns and roots, or else a 50 gl drop of a conidial suspension of FORL was put directly onto the crown of each seedling. In other experiments, seedlings were germinated in the dark on sterile Whatman #3 filter paper wetted with 0.5 ml of distilled water. These were incubated at 25 ° C. When the radicle was about 1 cm long the germlings were transferred to water agar plates that had been infested with the nonpathogen as above. These were left for another 3 days under lights, and then the seedlings were inoculated with the pathogen, as above.  Test Tube Assays  Either water agar or 1/4 strength Hoagland's solution agar (see Appendix B) was used to make slants in 16 x 100 mm or 25 x 200 mm test tubes (the larger tubes were used in the longterm timing experiments). Teardrop shaped pieces of glass microfibre (Whatman GF/A) or paper (Whatman #41) filters were laid over the slant to stop the plant roots from growing straight down into the agar. A 3-4 day old germling was placed at the top 30  of each slant and inoculated with 0.1 ml of the nonpathogenic conidial suspension, by pipetting the suspension onto the crown and roots. Unless otherwise specified, the plants were similarily inoculated with 0.1 ml of the pathogen conidial suspension 3 days after the first inoculation. Tubes were capped with translucent plastic Kap-uts. In other experiments, 2 ml of water or fertilizer solution were put into 16 x 100 mm test tubes. Four fertilizer solutions were used at different times (see recipes Appendix B): Hoagland's solution with 1/4 salt content (Hoagland and Arnon, 1938), foliar strength 20-20-20 (W.R. Grace and Co. Ltd., Ajax, Ont.), a fertilizer solution used by greenhouse tomato growers in the Fraser Valley (B.C. Ministry of Agric. and Fisheries, 1987), or a fertilizer solution designed especially for hydroponics in Holland (here called "Dutch Fertilizer Solution") (Sonneveld and Strayer, 1989). Wicks made from small (about 80 mm long) strips of glass microfibre (Whatman GF/A) or paper (Whatman #41) filters were bent in half and put into the tubes to draw up the liquid. Germlings (3-4 days old) were placed on top of the wicks, and inoculated with nonpathogenic strain and pathogen as in the first test tube experiments.  31  Comparison of Biological Control by B1 and Nit Bl, and Inoculum Density Experiments  The seedlings were grown in petri plates or test-tubes as noted in the results section. The inoculum densities and timing are noted in the results section as well, since this information is helpful for comparing results of different experiments. Day 0 always refers to seeding.  Timing Experiments  In the long-term experiment all the seedlings were inoculated at day 7 with either nit B1 or H 2 O (for the control). Some of the plants had already been inoculated with FORL on day 3, and some were inoculated with FORL on day 7, at the same time as nit B1. Other plants were inoculated at days 10, 17 or 28. The tubes were opened periodically, and all the seedlings were inoculated with 1 ml sterile fertilizer solution. In the short-term experiment all the seedlings were inoculated at day 3 with nit B1 or H 2 0. Some plants were also inoculated with FORL at day 3, while others were inoculated with FORL at days 4,5,6 or 7. Ratings in both timing experiments were taken at a constant time after the FORL inoculation (which meant that the ratings from each different timing treatment were actually taken at different days). 32  Seedling Growth Conditions  After the seeds germinated, the plants were kept either under a light bank in the lab, or else in a growth chamber. The photon flux density of the light bank varied from 61 moles 111 -2 s -1 to 114 Amoles m -2 s -1 , depending on the outside light coming through the window. The growth chamber had a photon flux density of 200 Amoles m -2 s -1 . Since the density readings were slightly different at different spots in the chamber, the petri plates or racks of tubes were periodically shifted around to ensure uniform growth conditions among them. Both the benchtop light bank and the growth chamber were set at a 16 h/8 h day/night photoperiod. The temperature in the lab ranged from 2025 ° C. The temperature in the growth chamber was set at a constant day/night temperature of 22 ° C, but it sometimes fluctuated between 18 ° C and 25 ° C. In all the experiments the plants grew much more slowly than they would have grown in soil, and their leaves were often chlorotic. Four different fertilizer solutions were tested in an attempt to improve plant growth, but none of them helped the plants grow any better (Fig. 1.2).  Disease Severity Ratings  At the end of each experiment plants were rated for disease severity on a scale of 0 to 4 (Fig. 1.3). 0 was a healthy plant, 33  Figure 1.2 Growth of tomato seedlings in different fertilizer solutions. Three week old seedlings (not inoculated) growing in 25 x 200 mm test tubes in the following solutions (L-R): Dutch fertilizer solution, filter paper wick; Dutch fertilizer solution, glass microfibre wick; Fraser Valley Greenhouse Grower's fertilizer solution, glass microfibre wick; Hoagland's solution agar, 1/4 strength salt content. (20-20-20 fertilizer solution not shown).  34  Figure 1.3. Disease severity rating scale for tomato seedlings inoculated with Fusarium oxysporum f. sp. radicis-lycopersici. From left to right: 0 (healthy), 1, 2, 3, 4, (5).  35  1 indicated a tiny brown speck on the crown, 2 was a larger speck or a region of slight browning at or near the crown, 3 was a long light brown lesion or a smaller dark brown lesion at the crown, and 4 was a large dark brown lesion at the crown. In the longterm timing experiment, a rating of 5 was used to distinguish plants that were actually dead from those with severe crown lesions. In other experiments, a dead plant was simply rated 4.  Statistical Analyses  Most of the experiments were factorial arrangements of completely randomized designs. In the petri plate experiments the mean disease severity rating of approximately ten seedlings per plate was calculated, and this mean was treated as an experimental unit. There were four replicate plates per treatment. In the test tube experiments each individual seedling was an experimental unit and there were usually ten seedlings per treatment. The main effects and interactions were partitioned into single degrees of freedom. In the simplest experiments, a significant interaction between the first inoculation (nonpathogenic strain or H 2 O) and the second inoculation (FORL or H2 O) was considered a positive indication of biological control. A significant three-way interaction between the two inoculations and some other factor meant that that factor had an effect on biological control. When disease severity ratings were taken twice, at different times, the experiment was analyzed as a 36  split-plot design, with the two rating dates as the two halves of the plot. In some experiments the least significant difference (a=0.05) was calculated so that means that were not separated in the ANOVA could be compared. ANOVA's are in Appendix A. When the standard error (S.E.) is reported it is 1M.S.  00 /n  where M.S.  (e)  is  the mean squared error and n is the sample size. Most of the experiments were repeated, although usually the methods were slightly different in each replication, leading to slightly different results. In most cases the results from both experiments are reported.  37  RESULTS  General Observations from Biological Control Experiments  Both nit B1 and B1 demonstrated consistent biological control of crown and root rot in the petri plate and test tube assays. Seedlings treated with B1 or nit B1 before being inoculated with Fusarium oxysporum f. sp. radicis-lvcopersici (FORL) were generally free of root rot at the end of each experiment, while plants treated with H 2 O before being inoculated with FORL were severely rotted or dead. When seeds were directly inoculated with B1 or nit B1 the fungal mycelium grew around the seed coat, holding it to the cotyledons, sometimes causing both seed coat and cotyledons to rot. As well, when the stems and leaves came in contact with this strain under high moisture conditions (for example, in the petri plate assays, when there was a lot of condensation on the lids of the plates), tiny water-soaked lesions developed, whether or not FORL was present (Fig. 1.4). If the environment remained wet, the lesions spread and the plants sometimes died. If the seedlings were pre-germinated on sterile filter paper and then transferred to petri plates infested with the nonpathogenic strain symptoms were much less severe, although it was still sometimes necessary to manually remove the seed coats. These symptoms rarely appeared on seedlings growing in test tubes where the stems and leaves did not contact free water. 38  Figure 1.4. Water-soaked stem and leaf lesions on tomato seedlings infested with biological control Fusarium strain nit B1  39  Comparison of Biological Control by B1 and Nit B1  The nit mutant was as effective as the wild-type strain at protecting the plants in both the petri plate and test tube assay (Table 1.2 and ANOVA, Table Al).  FORL Inoculum Density / Disease Severity Relationship  As  few as 10 conidia of FORL on the root gave rise to very  high disease severity ratings. Increasing the inoculum concentration beyond this level had little effect (Fig. 1.5). Regression analysis showed that a hyperbolic equation best described the relationship between the log concentration of FORL and disease severity ratings (see ANOVA, Table A2).  Parameters of Biological Control  1. Long-Term Timing  There was no significant difference (a=0.05) in the disease severity ratings of tomato seedlings inoculated with FORL 3 days, 10 days or 21 days after being inoculated with nit Bl (Fig. 1.6, and ANOVA, Table A3). Seedlings inoculated with FORL at the same time as nit B1 showed a significant delay in the onset of crown rot symptoms but this protection was not permanent - the plants became diseased and died after about a month. Seedlings inoculated with FORL before being inoculated with 40  ^  Table 1.2. Mean disease severity ratings of tomato seedlings inoculated with nitrate non-utilizing (nit) mutant or wild-type Fusarium strain B1 and then with Fusarium oxysporum f. sp. radicis-lycopersici (FORL) Disease severity a Inoculations  Experiment 1 b^Experiment 2 c  nit B1 + FORL  0.00^  0.5  wild-type B1 + FORL  0.00^  0.2  H 2 O FORL  1.5^  4.0  S.E.  0.22^ 0.12 (n=4)^ (n=10)  a On  a scale of 0 (healthy) to 4 (severe) Experiment 1: Seedlings were grown on 1% water agar. They were inoculated at seeding with H 2 O or 2.5 x 10 5 conidia of nit mutant or wild-type Fusarium strain, and inoculated 3 days later with approximatelyconidia of FORL (isolate 16). Ratings taken 1 week after FORL inoculation. `Experiment 2: Seedlings were grown on filter paper wicks in Dutch fertilizer solution. They were inoculated at day 6 with H 2 O or 2 x 10 2 conidia of nit mutant or wild-type Fusarium strain, and inoculated 3 days later with 8 x 10 3 conidia of FORL (isolate 16). Ratings taken 10 days after FORL inoculation. b  41  4  O  1  2  3  4  5  Log. FORL Conidiol Conc. Figure 1.5. Effect of Fusarium oxysporum f. sp. radicislycopersici (FORL) inoculum density on disease severity ratings of unprotected tomato seedlings. Seedlings were grown on slants of 1% water agar. They were inoculated at day 4 with a suspension of FORL conidia (isolate SSa 17-0 pink). They were rated on a scale of 0 (healthy) to 4 (severe) 1 week later. Each point represents the mean of 20 seedlings, pooled from 2 experiments.  42  Figure 1.6. Mean disease severity ratings, on a scale of 0 (healthy) to 5 (dead) of tomato seedlings inoculated with nonpathogenic Fusarium strain nit B1 and then with Fusarium oxysporum f. sp. radicis-lvcopersici (FORL) after increasing time intervals. Seedlings were grown in 25 x 100 mm test tubes on slants of 1/4 strength Hoagland's agar overlaid with GF/A filters. They were all inoculated at day 7 with H 2 0 or 2 x 10 3 conidia of nit Bl. Subgroups were inoculated at various times with 2 x 10 3 conidia of FORL (isolate SSc 33).  43  nit B1 were not protected. The long-term timing experiment was done twice with similar results; Figure 1.6 shows the mean disease severity ratings from the second experiment. An additional rating of "5" was used in this experiment to distinguish plants which were completely dead from those which had severe FORL lesions but were still living.  2. Inoculum Concentration  As few as 10 conidia of nit B1 (put on roots 3 days before the inoculation with 3 x 10 3 conidia of FORL) were sufficient to significantly reduce disease severity ratings relative to the FORL-only control (Fig. 1.7). Regression analysis showed that a hyperbolic relationship best described the relationship between the log concentration of nit B1 and disease severity ratings, in the presence of FORL (see ANOVA, Table A4).  3. Short-Term Timing  Plants inoculated with FORL 0, 1, 2, 3 or 4 days after being inoculated with nit B1 showed decreasing disease severity ratings (Fig. 1.8, and ANOVA, Table A5). There was a highly significant quadratic relationship between FORL inoculation timing and disease severity ratings, although this relationship was different at two different nit B1 inoculum densities. When the initial inoculation was 100 conidia of nit B1, the disease severity ratings decreased more quickly as the time interval before the FORL inoculation increased than when the initial 44  Log nit E0 Conidial Conc. Figure 1.7. Effect of nonpathogenic Fusarium strain nit B1  inoculum density on severity of crown and root rot in tomato seedlings induced by Fusarium oxysporum f. sp. radicislycopersici. Seedlings were grown on slants of 1% water agar. They were inoculated at day 4 with a suspension of nit B1 conidia, and at day 7 with 3 x 10 3 conidia of FORL (isolate SSc 33). They were rated on a scale of 0 (healthy) to 4 (severe) 2 weeks after the FORL inoculation. Each point represents the mean of 20 seedlings, pooled from 2 experiments.  45  Figure 1.8. Effect of increasing time intervals (from 0-4 days)  between inoculation with the nonpathogenic Fusarium strain nit B1 and inoculation with Fusarium oxvsporum f. sp. radicislycopersici on severity of crown and root rot in tomato seedlings. Seedlings were grown in test tubes in Dutch fertilizer solution. They were all inoculated on day 3 with H 2 0 or nit Bl, and, after various time intervals, with 6 x 10 3 conidia of FORL (isolate 16). The rating was on a scale of 0 (healthy) to 4 (severe) 46  inoculation was only 10 conidia of nit B1. This relationship changed slightly according to when the ratings were taken: as Fig. 1.8 shows, 10 conidia of nit B1 inoculated 2 or 3 days before FORL delayed the onset of disease symptoms, but did not provide permanent protection. In this experiment, the temperature in the growth chamber  °  increased to 27 C at one point, and all the plants looked stressed. Also, the FORL inoculum contained an unusually high proportion of macrospores. For whatever reason, the plants developed disease symptoms early, so the ratings were done after 1 week and 2 weeks instead of 2 and 4 weeks. The experiment was not repeated.  4.Nit Bi:FORL Ratio Biological control could be demonstrated when nit B1 and FORL were inoculated to the root simultaneously only if the concentration of nit B1 conidia was very high relative to the concentration of FORL conidia (Fig. 1.9, and ANOVA, Table A6). Disease severity ratings were taken at 10 days in the first experiment and at 20 days in the second experiment. In both, the ANOVA showed a highly significant linear relationship between the nit B1:FORL ratio and disease severity ratings.  47  Figure 1.9. Effect of different inoculum density ratios of  nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) on severity of crown and root rot symptoms in tomato seedlings. Experiment 1: Seedlings were grown in petri plates, and inoculated at day 3 with 1 x 10 2 conidia of FORL (isolate SSc 33) and 1 x 10 2 - 1 x 10 5 conidia of nit B1. Experiment 2: Seedlings were grown on slants of 1/4 strength Hoagland's agar, and inoculated at day 3 with nit B1 and FORL as in Experiment 1. Ratings were on a scale of 0 (healthy) to 4 (severe). 48  5. Absolute Number of Microconidia The absolute number of microconidia put on the plant (when the nit B1:FORL ratio was high enough to bring about biological control) did not substantially affect the disease severity ratings (Fig. 1.10, and ANOVA, Table A7). As in the previous experiments, the nit B1:FORL inoculum ratio had a significant overall effect. Higher concentrations of nit B1 relative to FORL gave rise to significantly lower disease severity ratings.  49  4 Nit 81 :FORL Ratio- 1001 0^ Nit B1 :FOR Ratio--1 0.1  0  10^100^1000 FORL hocarn Density (Conidio Plantl  Figure 1.10. Effect of total inoculum density (nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicislycopersici (FORL)) on severity of crown and root rot symptoms in tomato seedlings. Seedlings were grown in test tubes in Dutch fertilizer solution. They were inoculated day 10 with a mixture of nit B1 and FORL (isolate SSc 33) microconidia. Points represent the mean of 20 seedlings, pooled from two experiments, rated on a scale of 0 (healthy) to 4 (severe) 10 days after inoculation.  50  DISCUSSION  General Observations from Biological Control Experiments  Biological control was easily demonstrated in both the petri plate and test tube assays. The leaf and stem symptoms caused by the nonpathogenic strain in these experiments (Fig 1.4) have not been reported in any other experiments in which nonpathogenic strains of Fusarium oxysporum are used as biocontrol agents, even when seeds are inoculated. It is unlikely that these symptoms are important in tomato. They probably appeared as a result of the artificial conditions in the petri plates, in which the leaves and stems were often resting in drops of condensation on the petri plate lids. They probably would not appear under greenhouse conditions. However, F. oxysporum and F. solani have been associated with water-soaked lesions on the hypocotyls of soybean seedlings grown under conditions of high moisture, and these lesions can lead to significant crop losses (Farias and Griffin, 1990). It would be interesting (and important if Fusarium oxysporum strain B1 is to be used commercially) to compare the lesions found on tomato seedlings with those that occur on soybean seedlings. The leaf and stem lesions caused by the nonpathogen seldom appeared when plants were grown in test tubes, and the test tube assay is recommended for further research on this form of biological control.  51  Comparison of Biological Control by B1 and Nit B1  Nit B1 was as effective as its wild-type counterpart at protecting seedlings against crown and root rot, and there seemed no reason not to use it in place of the wild-type strain in subsequent experiments. The physiological phenotype of the mutant was examined and is explained in Appendix C.  FORL Inoculum Density / Disease Severity Relationship  The relationship between inoculum density of Fusarium oxysporum f. sp. radicis-lycopersici and disease severity ratings (Fig. 1.5) is difficult to compare to that found by other researchers studying Fusarium oxysporum diseases since most of these researchers used older plants and measured disease incidence rather than disease severity. Hartman and Fletcher (1991) found that a high incidence of seedling death (greater than 50%) in non-sterile rockwool cubes or peat blocks only occurred at high inoculum densities (10 6 microconidia per plant). In the system used here, however, a very small number of FORL microconidia (as few as 10 per plant) gave rise to severe disease symptoms. This is consistent with the infection process of FORL: it colonizes the root extensively from the point of infection, causing a rot that spreads from a small point, rather than 52  causing multiple small lesions. The small number of infectious propagules needed to cause disease makes simple competition a less likely mechanism of control in this system than in a system in which a higher number of infections are needed for disease to occur. As might be expected with this infection process, disease in all the experiments in this study was always an "all or nothing" phenomenon: plants rated "2" or "3" were in the process of developing severe disease symptoms and always reached a rating of "4" if they were left for another week. The advantage of keeping the rating system, rather than simply recording disease incidence was that the rating system allowed detection of the delay in the onset of disease that was observed in some treatments.  Parameters of Biological Control  1. Long-term Timing Biological control by nonpathogenic strain nit B1 was effective when the interval between inoculations was as long as 21 days (Fig. 1.6) and this control persisted for up to 48 days (this was the longest that any experiment lasted). If competition is the mechanism of protection, then, under the conditions of this experiment, isolate nit 81 is a good root colonizer, capable of maintaining a constant presence on all parts of the growing root that might otherwise be colonized by the pathogen. If induction of a transitory host defense response is the mechanism 53  of protection, then again nit B1 must be a good root colonizer, for it must constantly re-induce a response in the growing roots. The induction of a long-term host resistance response in the plant is a possible mechanism of protection that would not depend on nit B1 persisting on all parts of the roots. However, in most reports of induced resistance against Fusarium oxysporum pathogens, the control is transitory (Wymore and Baker, 1982; Komada, 1990) and is sometimes less effective when there is a long delay between inoculation with the inducer (here, the nonpathogenic Fusarium) and inoculation with the challenger (here, FORL) (Wymore and Baker, 1982; Matta, 1989; Jorge et al., 1992), suggesting that the induced resistance reaction in most plants is generally short-term.  2. Nit B1 Inoculum Concentration and Short-Term Timing  The proportion of nit B1 microconidia to FORL microconidia that was needed to give good control was inversely related to the interval between inoculations. When the interval between inoculation with nit B1 and inoculation with FORL was 3 or 4 days, just 100 microconidia of nit B1 on the roots provided good protection even when plants were inoculated with 3000 or 6000 microconidia of FORL (Figs. 1.7 and 1.8 respectively). As the interval between inoculation with nit B1 and inoculation with FORL decreased, the protection brought about by a small initial inoculum declined (Fig. 1.8). When the two strains were inoculated simultaneously the concentration of nit B1 conidia 54  relative to FORL conidia had to be 10:1 for there to be any protection and 100:1 for complete protection (Fig. 1.9). These results can be explained by any of the four mechanisms of biological control described in the literature review. If infection sites, or substrates necessary for infection are very limited, then prior inoculation with enough nit B1 to take up all those sites or substrates might allow biological control by exclusion of the pathogen. For this hypothesis to be possible the maximum number of infection sites on the root must be extremely low. In fact, this maximum could be as low as 10, for in some cases roots inoculated with just 10 microconidia of nit B1 were completely protected. This seems unlikely, especially when one considers the fact that not all of the microconidial suspension used for the inoculation necessarily hits the roots. However, maximimum colonization rates of healthy roots by other Fusarium  oxysporum strains reported in the literature are indeed very low. The number of colony forming units that can be isolated from 100 cm of root, even after inoculation with a much higher density of microconidia, rarely exceeds 100 (Taylor and Parkinson, 1961; Gordon et al., 1989; Mandeel and Baker, 1991). This number does not accurately reflect the actual amount of fungus, either as hyphae or as spores, present on the root, since in most of these experiments the root is cut into a finite number of segments and the number of segments with hyphae growing out is reported. What the number shows, rather, is that there are segments of the root at any given time that, for some reason, remain uncolonized. This 55  idea is supported by microscopic observations (Foster et al., 1983). Given this idea, and keeping in mind the fact that nit B1 has time to increase on the root in the interval between inoculations, the competition hypothesis should at least be considered. With simultaneous inoculations, the two fungi would be competing on an equal footing, and whichever isolate was present in greater numbers would be successful in colonizing the root. Competition is often suggested as a mechanism in experiments dealing with biological control by nonpathogenic  Fusarium oxysporum strains in suppressive soils. In these experiments the nonpathogenic strains are inoculated simultaneously and at least a 10:1 ratio between nonpathogen and pathogen is needed for significant control (Paulitz et al., 1987; Lemanceau et al., 1992). Another explanation is that nit B1 stimulates a systemic resistance response in the plant. The induction of systemic defense responses in plants is still poorly understood, but one model, which has been proposed for tomato as well as other plants, is that a signal molecule, possibly of plant origin, is translocated through the plant from the point of infection and "primes" all the cells in the plant to respond quickly to subsequent attack (Kuc, 1990; Van Peer and Schippers, 1992). The need for an interval between inoculations would reflect the need for time for the signal molecule to spread though the whole root system before protection was complete. The effect of nit B1 inoculum density, even after the lag period, might reflect the 56  fact that a greater density of nit B1 stimulates the release of a greater amount of signal molecule. Nit B1 could also stimulate a localized defense response in the plant, one that again takes time to take effect, but whose spread is caused in part by the spread of the inducing organism. It has been shown that at least some nonpathogenic fusaria are capable of colonizing the cortex of young tomato roots and these fusaria seem to stimulate a defense response that prevents them from destroying this tissue (Bishop and Cooper, 1983). The response is generally seen 2 or 3 days after inoculation and is limited to cells directly adjacent to fungal hyphae. Cells are bigger than microconidia. If each microconidium that sticks to the root induces a change in a whole cell then the effect of a limited number of conidia, once the cell changes have occurred, is magnified. A larger proportion of protecting microconidia would be required when plants are inoculated simultaneously with the two strains since the cell changes would not yet have occurred and simple competition for unmodified colonization sites or for substrates would then be more important. A fourth possible explanation is that nit B1 produces some antibiotic that directly inhibits the growth of FORL or else induces the plant to produce a phytoalexin to which it (nit B1) is immune. This explanation has been suggested in reports of biological control of Fusarium wilt of sweet potato by a nonpathogenic Fusarium oxysporum (Komada, 1990).  57  3. Absolute Number of Microconidia  Changing the absolute number of microconidia placed simultaneously on the root had no effect on biological control (Fig. 1.10). If some direct interaction like antibiosis is important when nit B1 and FORL are inoculated simultaneously, then one might expect that there would be less protection when the absolute inoculum density is lower since the distance between a pathogenic microconidium and a nonpathogenic microconidium would be greater. With a low total inoculum density it might be possible for a germinating microconidium of FORL to penetrate the plant before encountering the influence of nit B1. Likewise, if cell changes induced by the nonpathogen were indeed very localized then a solitary FORL hypha that somehow got beyond the changed cells would also be able to cause disease. However, if the nonpathogenic strain induces a host resistance response that affects all of the root cells, or if this strain spreads in a limited fashion though the root cortex, inducing changes as it goes, then a solitary FORL hypha will be very limited in the extent of damage it can cause regardless of total initial inoculum density. Although the FORL hypha may start out far from nit B1 it will sooner or later encounter that strain or the effects of that strain. The competition hypothesis depends on there being a very limited number of infection sites on the root. The lowest total number of microconidia used in this experiment (100 of nit B1 and 10 of FORL) would be great enough to saturate those sites so 58  increasing the total number of microconidia over that number would have no further effect on biological control.  59  Chapter 2  Colonization of Tomato Seedling Roots by Nonpathogenic Fusarium Strain Nit B1 and  Fusarium oxysporum f. sp. radicis-lycopersici  INTRODUCTION  Experiments in Chapter 1 showed that the ratio of nonpathogenic Fusarium strain nit B1 to Fusarium oxysporum f. sp. radicis-lycopersici (FORL) had to be at least 10:1 for protection to occur when the two fungi were inoculated to the roots simultaneously. However, a lower inoculum density of nit B1 was successful when inoculation with the pathogen was delayed. None of the four possible mechanisms of biological control were disproved in Chapter 1, but competition, for nutrients or for infection sites seemed unlikely, especially in the case of biological control by a small number of nit B1 microconidia inoculated to the roots 3 days before FORL. Competition could only be a mechanism in that case if infection sites or substrates necessary for infection were extremely limited, so that 10 microconidia of nit B1 could grow enough to take up all those sites after 3 days. The colonization experiments were designed to further test competition as a mechanism of biological control. There are as yet no methods for measuring the actual number of fungal 60  infections on a root. The number of spores of a given strain can be measured, by taking a sample of water in which the roots have been rinsed, but this number may not reflect the number of established colonies of that strain. A better method is to grind the roots and count the number of fragments that have any fungal colonies growing from them. The problem with simply grinding the root, however, is that the source of these colonies can be spores, hyphae growing along the root surface, or hyphae that have actually penetrated the plant and caused infection. Researchers often soak root sections in a 0.1% (Farias and Griffin, 1990) or 0.21% (Paulitz et al., 1987; Schneider, 1984) aqueous solution of NaC1O for up to 3 minutes to surface sterilize them before recovering Fusarium oxysporum strains. In preliminary experiments to the present research, however, virtually no colony forming units of nit B1 were recovered after the roots were soaked in a 0.21% aqueous solution of NaC1O for 2 minutes (data not shown), so no bleaching step was included. Taylor and Parkinson (1961) recommended rinsing root samples in 20 changes of sterile distilled water to remove fungal material loosely associated with the root surface. That procedure was considered to be too time-consuming for the present research, so instead the roots were rinsed vigorously in three changes of distilled water before being ground. The colony forming units obtained by this method probably come from hyphae growing on the surface and within the root, rather than from spores. Colonies of nit B1 that have become well-established on the root surface, but 61  have not necessarily penetrated the root, may play a role in limiting root colonization by FORL. Although the recovery procedure used in this experiment may not reflect actual infections by either nit B1 or FORL, it should reflect root colonization accurately enough to test the competition hypothesis. If competition is the mechanism of biological control there should be a correlation between root colonization by each strain and biological control, with long-term control occuring only when colonization by nit B1 is at a maximum and colonization by FORL is at a minimum. As well, root colonization by nit B1 should increase rapidly in the time between inoculations, reaching a maximum after 3 or 4 days. This would support the idea that there is a limited number of infection sites on the root and that these sites can be filled after 3 or 4 days by nit Bl. The objective of the preliminary colonization experiments in this chapter was to see if there was any correlation between root colonization by nit B1 and FORL, and biological control. The objectives of the short-term and long-term timing experiments were to determine the rate of increase of nit B1 in the interval between inoculations, and to see if root colonization ever reached a plateau, reflecting a limited number of infection sites.  62  METHODS  Selective Media  Fusarium minimal medium (FMM) and chlorate-minimal medium (FMM-C) as described in Chapter 1 were used to monitor populations of Fusarium oxysporum f. sp. radicis-lvcopersici (FORL) and nonpathogenic Fusarium strain nit B1 on tomato roots. A preliminary experiment showed that there was no difference in the percentage germination of FORL microconidia on FMM and the pergentage germination of nit B1 microconidia on FMM-C (data not shown). Maximum colony counts on FMM were obtained after 5 days and maximum colony counts on FMM-C were obtained after 7 days.  Preliminary Colonization Experiments  Seedlings were grown in test tubes on slants of 1% water agar overlaid with tear-drop shaped pieces of glass microfibre filter paper (Whatman GF/A) as described in Chapter 1. They were inoculated at various times with suspensions of nit B1 or FORL microconidia at the inoculum densities noted in the results section. After varying incubation periods, the seedlings were taken off their slants. A 20-mm segment (5 mm above the crown and 15 mm below) was aseptically cut from each seedling, rinsed three times in 5 ml of sterile distilled water in a test tube by vigorously swirling on a vortex blender, then transferred to a 63  small mortar containing 1 ml sterile distilled water. The root segment was ground to a fine slurry using a sterile pestle. Part of the slurry (0.4 ml) was spread onto FMM and part (0.4 ml) was spread onto FMM-C.  Short-term Colonization Experiment  Seedlings were grown on glass microfibre wicks in Dutch fertilizer solution as described in Chapter 1. They were inoculated at day 4 with 0.1 ml of a suspension of nit B1 microconidia, and at day 11 with 0.2 ml of a suspension of FORL microconidia made up in Dutch fertilizer solution. Colony forming units were recovered daily until day 14. Seedlings (10 each day) were taken off their wicks and a 20-mm segment (5mm above the crown and 15mm below) was aseptically cut from each seedling. The roots were combined in groups of two (each day there were five replicate groups). The two segments were transferred to a small mortar, and were ground together to a fine slurry in 2 ml sterile distilled water, using a sterile pestle. Part of the slurry (0.2 ml) was spread onto FMM and part (0.2 ml) onto FMM-C.  Long-term Colonization Experiment  Seedlings were grown on paper wicks in Dutch fertilizer solution as described in Chapter 1. They were inoculated at day 4 with 0.1 ml of a suspension of nit B1 microconidia and at day 8 64  with a suspension of FORL microconidia. Colony forming units were recovered daily until day 20. Each day, 15 seedlings were taken off their wicks. The secondary roots were aseptically removed and a 20-mm segment was cut from each tap root as described above. These segments were combined in groups of three (each day there were were five replicate groups) and rinsed in 3 ml sterile distilled water by vigorously swirling in a test tube on a vortex blender. Five ml of the rinse water was spread onto FMM and 5 ml onto FMM-C. Then the roots were separated and each root was rinsed individually three times in 5 ml sterile distilled water, again by vigorously swirling in a test tube. Each root was ground alone in 1 ml sterile distilled water. Then the slurries from the three roots in each group were combined. Part of the slurry (0.5 ml) was spread onto FMM and part (0.5 ml) onto FMM-C.  Statistical Analyses  In the preliminary experiments colony forming units (CFU's) of nit B1 and FORL were compared using an F-test (Experiment 2 was treated as a 2 x 2 factorial experiment). Regression analysis was performed on the data from the short- and long-term colonization experiments, and in the long-term experiment a ttest was used to compare slope parameters (Steel and Torrie, 1980). ANOVAs are in Appendix A. The colony counts from all the experiments are reported as CFU's per 20 mm root segment to facilitate comparisons. 65  RESULTS  Preliminary Colonization Experiments  There was no correlation between colony counts of nit B1 and FORL, and biological control (Table 2.1). The numbers of colony forming units (CFU's) of each type recovered at the end of each experiment seemed to be related to the initial inoculum concentration. In Experiment 1, in which the initial inoculum density of FORL was higher than that of nit Bl, colony counts of FORL at the end of the experiment were significantly higher (a=0.01) than colony counts of nit B1 (see ANOVA, Table A8). These colony counts were still much lower than those from the FORL-only controls. In Experiments 2 and 3, in which the initial inoculum densities of nit B1 and FORL were approximately the same, there were no significant differences between final colony counts of each strain at the end of the experiment. Disease symptoms developed in the FORL-only controls but not in any other plants in these experiments.  66  Table 2.1. Colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) recovered from previously-inoculated tomato seedling roots Inoculum density^ CFU's per 2 cm root segment Disease (microconidia^Inoculum^ per root)^timing^ nit B1^FORL^ Symptoms  Experiment 1 nit Bl: 10^-nit B1 at seeding, ^0^254^S.E.=43^no FORL: 8x10 4^FORL at day 4^ (n=8) -grinding at day 7 no nit Bl: 5x10 3^2 ^224^ FORL: 8x10 4 (FORL-only control)^  Experiment 2 (/` nit Bl: 1x10 5^-nit B1 at seeding, FORL: 1x10 5^FORL at day 4 -grinding at day 10  a  15^11^S.E.=4.6 (n=8) ++8  (FORL-only control)  Experiment 3 nit Bl: 2.5x10 4 FORL: 1x10 4  877^  -nit B1 at day 4, FORL at day 8 -grinding at day 16  Too many colonies to count  168^97^S.E.=38 (n=8)  yes  no  yes  no  Short-term Timing Experiment  The number of colony forming units (CFU's) of nit B1 recovered from unrinsed roots initially remained constant (from day 8 to day 10) and then increased (Fig. 2.1). The number of CFU's of FORL (isolate SSc 33) remained constant over the course of the experiment. A linear equation best described the increase of CFU's over time (see ANOVA, Table A9). No plateau in counts of CFU's was reached, and no disease developed over the course of this experiment. Colony forming units counted in this experiment probably arose from spores, as well as from hyphae on the root surface and in the root, since there was no rinsing step.  Long-term Timing Experiment  Colony forming units of nit B1 rinsed from the surface of the root (reflecting sporulation) increased approximately 100fold in the interval between the two inoculations (Fig. 2.2 (A)). Colony forming units of FORL (isolate 16) remained very low from day 8 (inoculation) to day 10, and then began to increase. A linear equation best described the increase of colony forming units of nit B1 over time, while a cubic equation best described the change in colony forming units of FORL (see ANOVA, Table A10). Colony forming units of nit B1 closely associated with the root surface or inside the roots increased approximately 50-fold 68  2200 2000 1800 a  1600  cE 1400 g 1200  1000 800 600 400 200  0  8^9^10^11^12  13  14  Time (Days Post Seeding)  Figure 2.1. Colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) recovered from unrinsed tomato roots over 10 days. Roots were inoculated with 3 x 10 2 conidia of nit B1 on day 4. They were inoculated with 1 x 10 4 conidia of FORL (isolate SSc 33) on day 11. Each point represents the mean of 5 samples.  69  Figure 2.2. Colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici washed from the surface (A) or recovered from rinsed, ground tomato roots (B) over 20 days. Seedlings were inoculated with 2.5 x 10 2 microconidia of nit B1 on day 4 and 1.5 x 10 3 microconidia of FORL (isolate 16) on day 8. Each point represents the mean of 5 samples. 70  in the interval between inoculations (Fig. 2.2 (B)). Colony forming units of FORL again remained constant until day 10, and then began to increase. A cubic equation best described the change in CFU's of nit B1 while a linear equation best described the increase in FORL CFU's (see ANOVA, Table A10). The linear components of the increasing portions of the graphs were compared using a t-test (Steel and Torrie, 1980). The rate of increase of nit B1 CFU's was significantly higher (a = 0.05) than the rate of increase of FORL CFU's. No plateau in colony counts of either nit B1 or FORL was reached over the course of the experiment. No plants that were ground up in this experiment had any symptoms of disease. However, more plants were inoculated for the experiment than were actually ground up, and at the end of 20 days many of the remaining plants showed signs of crown and root rot, and some of them had died, indicating that biological control was beginning to break down.  71  DISCUSSION  Preliminary Colonization Experiments  Competition was not demonstrated in any of these experiments. In the first experiment virtually no colony forming units of nit B1 were detected while more than 200 of FORL were recovered, yet the plants were still healthy at the end of the experiment. These plants were ground up only 4 days after the FORL inoculation, so they might have developed disease had they been left longer. However, the onset of disease was at least delayed, since the FORL-only controls included in this experiment developed some disease symptoms after just 4 days. In Experiments 2 and 3 the roots were ground up 6 and 8 days after the FORL inoculation, and still no disease symptoms developed. Similar results have been found in other systems. Mandeel and Baker (1991) found that two nonpathogenic isolates of Fusarium oxvsporum gave control against Fusarium wilt of cucumber, yet only one of them interfered with root colonization by Fusarium oxysporum f. sp. cucumerinum. Schneider (1984) tested a number of nonpathogenic isolates of Fusarium oxysporum for their ability to control Fusarium yellows of celery. He found that the isolate that gave the best control was the isolate that reduced colonization by F. o. apii the most. However, other nonpathogenic isolates gave some biological control, as measured by increased root dry weight and a reduced disease severity index 72  even though they did not affect the pathogen root colonization rates at all. These authors concluded that competition for infection sites might be a major factor in biological control by nonpathogenic fusaria. However, the fact that the complete prevention of infection by the pathogen is not essential for that control suggests that once the nonpathogenic strains have colonized the roots in sufficient numbers they may also affect the susceptibilty of the whole root (all possible infection sites) to disease. The proportions of CFU's of each type recovered from protected plants at the end of each experiment seemed to be related to the proportions of microconidia put on, despite the lag in inoculations. This suggests that nit B1 does not increase substantially in the interval between inoculations, perhaps not enough to overtake FORL when FORL is inoculated. These results support the idea that both the nonpathogen and the pathogen have equal abilities to persist in or on colonization sites on the tomato root, and the ability for one to exist in greater quantities than the other is more a reflection of the fact that it was put on in greater numbers than of any inherent competitive advantage that it may have. Similar results have been found before (Schneider, 1984; Mandeel and Baker, 1991). The idea that plant pathogens, especially Fusarium oxysporum pathogens, may be good root colonizers with a high competitive ability even on healthy roots has only recently become widely accepted (Lockwood, 1988). Previously, the competitive ability of a fungus was 73  thought to be inversely related to its degree of pathogenic specialization, and Fusarium wilt pathogens, with a high degree of specialization coupled with a small number of points of infection, were thought to be poor competitors (Garrett, 1979). It is still agreed that Fusarium oxysporum pathogens are poor competitors outside the rhizosphere (Banihashemi and deZeeuw, 1975; Marois et al., 1981; Gordon et al., 1989), so they could still be said to have a poor competitive saprophytic ability. However, more and more evidence is showing that when only root surface or root cortex colonization is measured, pathogenic strains are as common as nonpathogenic strains, no matter what the host (Parkinson, 1967; Gordon et al., 1989).  Timing Experiments  The results from the two timing experiments, using two different isolates of FORL, were somewhat similar (Figs. 2.1 and 2.2). In both, colony forming units of nit B1 increased significantly in the interval between inoculations, so that at the time of the FORL inoculation at least 50 times more CFU's of nit B1 than of FORL were recovered. It seems that the lag in inoculations may be important in giving nit B1 time to increase on the root before FORL is introduced. The discrepancy between this result and that of the first two preliminary experiments can perhaps be explained by the fact that in the timing experiments the seedlings were inoculated on day 4, when the seedling roots 74  were about 1 cm long, while in the first two preliminary experiments the seeds themselves were inoculated. In the latter case, the inoculum potential of nit B1 may have decreased in the interval before the root, a rich substrate for growth, grew down the water agar slant. This result does not mean that competition is necessarily involved in biological control. For competition to be a mechanism, the number of infection sites, or else the amount of substrate necessary for infection, must be limited. CFU's recovered from rinsed roots might be expected to reach a plateau if this is the case (CFU's from unrinsed roots in Figures 2.1 and 2.2 (A) may reflect sporulation, which could continue even after the root was fully colonized). As Figure 2.2 (B) shows, such a plateau was never reached. It is possible, however, that the CFU's in Figure 2.2 (B) come from hyphae closely associated with the root surface but not inside the root. This surface colonization might continue even after all possible infection sites are filled. Thus it cannot be stated with certainty that infection sites are unlimited in this system. In both timing experiments, the rate of increase of nit B1 was greater than the rate of increase of FORL. This greater rate of increase could be due to the fact that nit B1 was put on first, or it could mean that nit B1 has an intrinsically greater ability to colonize intact roots. The distinction might be tested by following an inoculum procedure such as that used in the first two preliminary experiments, in which counts of nit B1 CFU's at 75  the time of the FORL inoculation were very low. If nit B1 has some advantage other than the fact that it is put on the roots first then colonization by nit B1 might be expected to overtake colonization by FORL. There were no FORL-only controls in the timing experiment. However, given the results of the FORL-only controls in the preliminary experiments, it would be logical to assume that CFU's of FORL would increase at an extremely fast rate in the absence of nit Bl. Root colonization by FORL is therefore restricted in the presence of nit Bl. This is consistent with mechanisms of either competition or the elicitation of a host defense response. If antibiosis were involved, however, colonization by FORL would probably decrease over time.  76  Chapter 3  Testing Induced Resistance as a Possible Mechanism of Biological Control  INTRODUCTION  The "best" explanation for the biological control that was demonstrated in Chapters 1 and 2 was that nonpathogenic strain nit B1 induced a defense response in the tomato roots that made them resistant to subsequent infection by FORL. This mechanism seemed more likely than competition, because a very small number of nit B1 microconidia inoculated to the root could protect the plants against a much higher concentration of FORL inoculum, as long as there was a sufficient interval between inoculations. Although nit B1 probably did increase on the root during this interval, the increase in colonization (estimated by the increase in colony forming units that could be recovered from ground roots) was not sufficient to explain the biological control by competition alone. As well, biological control occurred even when colonization by nit B1 was no greater than colonization by FORL. The mechanism of induced resistance was tested more directly in this chapter. Much of the research on resistance in tomato has involved a pathogen and a resistant host cultivar. Thus Brammal and Higgins 77  (1988) studied the response of the resistant cultivar Ohio MR13 to invasion by FORL. They observed that cortical cell wall alterations, such as wall thickening and the formation of papillae, were successful in limiting the pathogen to the hypodermis of the root. Related biochemical studies have shown that fungal cell wall-degrading enzymes, notably B-1,3 glucanases and chitinases, accumulate in tomato cultivars resistant to FORL (Benhamou et al., 1989; Benhamou, Joosten and De Wit, 1990). These enzymes may directly destroy the invading fungal hyphae, and they may also release fragments from the fungal cell wall which elicit later defense responses (Benhamou, Joosten and De Wit, 1990). B-1,3 glucans of fungal wall origin have been shown to elicit such defense responses as the accumulation of antimicrobial phytoalexins (Ayers et al., 1976) and the production of ethylene and the accumulation of hydroxyproline-rich glycoproteins (HRGP's) (Roby et al., 1985). HRGP's may be precursors to the formation of structural barriers in tomato cortical cells (Benhamou, Mazau and Esquerre-Tugay6, 1990). The same defense responses occur in susceptible hosts, but they occur later, after the fungus has already caused extensive degradation of the cortex (Brammal and Higgins, 1988; Benhamou, Joosten and De Wit, 1990). Similarily, researchers studying the responses of tomato roots to Fusarium oxysporum f. sp. lycopersici (FOL), the tomato wilt pathogen, have suggested that the responses that limit the spread of this pathogen in vessel elements occur more quickly in resistant cultivars than in susceptible cultivars 78  (Bishop and Cooper, 1983). The reaction that is important in the present research, if induced resistance is the mechanism of biological control, is not a response of a particular resistant host cultivar. Rather, the response that is important here is the cortical response of any host, susceptible or resistant, to colonization by either a nonpathogenic Fusarium oxysporum strain or a wilt pathogen. This response is noted in reports of wilt resistance, reports that generally go on to discuss in depth the later vascular response to colonization by these strains (Beckman and Talboys, 1981; Bishop and Cooper, 1983). The response is similar, at least structurally, to the response that occurs in FORL-resistant tomato roots: there is an activation of cytoplasm in the vicinity of fungal hyphae, a thickening of cortical cell walls, and a formation of wall appositions such as papillae (Bishop and Cooper, 1983). This response does not restrict the invading fungal hyphae to the hypodermis. Instead, it seems to limit the invasion of cortical cells so that the hyphae grow towards the vascular cylinder between the cortical cells, rather than within them (Beckman and Talboys, 1981; Bishop and Cooper, 1983). Since this response is closely associated with the contact of fungal hyphae and the host cortical cells (Bishop and Cooper, 1983), it may be triggered by a fungal cell wall component. And since the response is, at least structurally, similar to responses that occur in resistant hosts when challenged with a pathogen, the same components that elicit responses in resistant 79  hosts, for example the B-1,3 glucan, may elicit this response. Crude fungal cell wall elicitors can be obtained by autoclaving whole culture mycelium or culture filtrates (Roby et al., 1985) or by grinding and centrifuging fungal mycelium to separate the cell wall from the cytoplasm (Ayers et al., 1976). In most examples of biological control of Fusarium oxysporum pathogens by induced resistance only a living inoculum of the avirulent or nonpathogenic strain can bring about the biological control (Matta, 1989). However, Mas et al. (1981) found that a filtrate from an avirulent strain of F. o. melonis brought about an attenuation of disease symptoms in musk-melons infected with the virulent strain of that pathogen. The objective of this chapter was to see whether non-living elicitor preparations of nit B1 could, similarily, bring about biological control in this system.  80  METHODS  General Procedures  Tomato seedlings were grown on water agar in Petri plates or test tubes, as described in Chapter 1. Seedlings were inoculated with a potential elicitor at various times prior to and after inoculation with 0.1 ml of a FORL microconidial suspension. In most cases the elicitor preparation was simply pipetted onto the roots and crowns, but for the heat-released cell wall fragment experiment the crowns were pierced with a sterile pin before the preparation was put on, to allow the preparation to reach cells below the epidermis. Plants were rated for disease severity 1 or 2 weeks after the FORL inoculation. The elicitor preparations are described below.  1. Elicitors from Fungus Grown in Nutrient Broth  A small piece of mycelium of Fusarium strain nit B1 was put into 100 ml of potato dextrose broth (PDB) or Fusarium minimal broth consisting of Fusarium minimal medium (FMM without agar: see Appendix B) supplemented with 2 g 1 -1 L-asparagine as a nitrogen source. After 5 days of growth in shake culture at room temperature, the contents of the flasks were strained through cheesecloth and then filtered, with the addition of water, through a 0.45 Am nitrocellulose filter (Millipore, GS). The filtrates were condensed to about 50% original volume using a 81  Roto-evaporator (Rotovapor, Brinkman) at 40 ° C and half of each condensed filtrate was dialyzed against distilled water for 24 hours at room temperature. All solutions were filter-sterilized through 0.22 Am filters prior to use. A suspension of heat-killed spores was made by centrifuging the broth at 5,000 rpm for 10 minutes and resuspending the pellet in distilled water to give a concentration of approximately 4 x 10 6 microconidia m1 -1 , which was autoclaved for 5 minutes (liquid cycle). In a second experiment only filtrates from cultures growing in potato dextrose broth were used, and no additional water was added at the first filtering step. Filtrates were condensed to 50% or 25% original volume. After dialyzing, filtrates were filter-sterilized through 0.45 Am filters.  2. Elicitors from Nit B1-Infested Germlings  About 0.7 g of surface-sterilized tomato seeds were put into 40 ml sterile distilled water in 250 ml Erlenmeyer flasks. A suspension of nit B1 microconidia was added to half the flasks to give a concentration of approximately 1.5 x 10 6 microconidia m1 1 . Flasks were incubated under regular laboratory lights at room temperature on a rotary shaker (100 rpm). After 6 days the liquid in the flasks was strained through two pieces of Whatman #1 filter paper and then filtered through a 0.45 Am nitrocellulose filter. The following elicitor preparations were made: 1) filtrate: the filtrate was condensed and dialyzed as in the first experiments, and filter sterilized through a 0.22 Am 82  nitrocellulose filter prior to use; 2) heat-killed mycelium/spore suspension: the extract from the 0.45 gm filter was re-suspended in sterile distilled water and autoclaved for 5 minutes; and 3) control: filtrate from non-infested (sterile) germlings.  3. Elictors fron Nit Bi-Infested Seedlings  About 250 surface-sterilized tomato seeds were put into a deep culture dish with enough sterile distilled water to barely cover the bottom of the dish. The seedlings were grown for 10 days, with sterile distilled water added as necessary. Then they were inoculated with 20 ml of a turbid suspension of nit B1 microconidia. After one week four different preparations were made: 1) root washings: these were made from the liquid from the bottom of the dish, drawn off and filtered through a 0.45 gm nitrocellulose filter; 2) root and stem slurry: the leaves were cut off and the roots and stems were ground in a mortar and pestle, autoclaved for 5 minutes, and then strained through 4 layers of sterile cheesecloth; 3) root and stem supernatant: some of the root and stem slurry was centrifuged at 11,000 rpm for 15 minutes, and then the supernatant was decanted off and filter sterilized; 4) positive control: this consisted of a ground stem and root slurry that was not sterilized. All the elicitor preparations were stored at 4°C before use.  83  4. Heat-Released Fungal Cell Wall Fraction  Both nit B1 and FORL were grown in Potato Dextrose Broth as in the first experiment. After 2 weeks, thick mycelial mats were harvested by filtering through glass microfibre filters (Whatman GF/A), and rinsing with sterile distilled water. The mycelium was resuspended in sterile distilled water (5 g per 100 ml) and blended in a sterile Virtis blender. The slurry was autoclaved for 5 minutes, then centrifuged at 11,000 rpm for 15 minutes. The pellet and supernatant were separated and both were stored at 4 ° C before use.  Statistical Analyses  As in Chapters 1 and 2 the experiments were factorial arrangements of completely randomized designs. They were subjected to ANOVA, using the L.S.D. test to separate means where necessary.  84  RESULTS  None of the nit B1 broth filtrates, the exudates from nit B1-infested germinating seeds or older seedling roots, the heatkilled spores or the fungal cell wall fraction protected the plants against disease caused by FORL (Tables 3.1 to 3.4, and ANOVAs, Tables All to A13). Plants treated with sterile exudates from nit B1-infested roots (Table 3.3) and with sterile cell wall debris (the pellet fraction) from both nit B1 and FORL (Table 3.4) had significantly higher disease severity ratings than plants treated only with H 2 O before being inoculated with FORL. In the former, the elicitor preparation, which contained compounds from ground tomato roots, may have been a source of extra nutrition for the pathogen. In the latter, the elicitor preparation, a cell wall fraction, was thick and dense and may have interfered with root aeration.  85  Table 3.1. Mean disease severity ratings of plants treated with possible elicitors from broth cultures of Fusarium strain nit B1 and then inoculated with Fusarium oxysporum f. sp. radicislycopersici (FORL) Disease severity a  Elicitor preparation  no FORL^with FORL  Experiment 1 Filtrate from: Potato dextrose broth, undialyzed (condensed 50%) Potato dextrose broth, dialyzed (condensed 50%) Fusarium minimal broth, dialyzed (condensed 50%)  0.0  4.0  0.1  3.6  0.0  3.6  Heat-killed spores  0.1  3.6  H2 O  0.0  3.6  Living nit B1 (1 x 10 6 conidia m1 -1 )  0.3  0.8  S.E.^=^0.17^(n=4) Experiment 2 Filtrate from: Potato dextrose broth, undialyzed -original volume -condensed 50% Potato dextrose broth, -original volume -condensed 50% -condensed 75%  0. 0 0. 1  4.0 4.0  0.1 0.0 0.1  4.0 4.0 4.0  H2 O  0. 0  4.0  Living nit B1 (1 x 10 6 conidia m1 -1 )  0.1  0.0  (Experiment not analyzed; n=10) a sOn  a scale of 0 (healthy) to 4 (severe), rated 10 days after the FORL inoculation Experiment 1: Seedlings were grown on plates of 1% water agar. They were treated with a drop of filtrate solution at seeding and at day 4. A mycelial plug of FORL (ssC 33) was put in the center of the plates at day 4 (nothing put in center of control plates). Experiment 2: Seedlings were grown on slants of 1% water agar in test tubes. They were treated with a filtrate solution at days 4,5,6 and 7, and with H 2 O or 3.5 x 10 2 conidia of FORL on day 8. 86  Table 3.2. Mean disease severity ratings of plants treated with possible elicitors from nit B1-infested germlings and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL) Elicitor preparation^  Disease severity a  Exudate from nit Bi-infested germlings: ^ -not condensed^ -condensed 50% ^ -condensed 50%, dialyzed ^ -condensed 75% ^ -condensed 75%, dialyzed Exudate from sterile germlings: ^ -not condensed^ -condensed 50% ^ -condensed 50%, dialyzed ^ -condensed 75% ^ -condensed 75%, dialyzed ^ Heat-killed spores H2 O  ^  Living nit B1 (5 x 10 4 conidia m1 -1  4.0 4.0 3.9 4.0 4.0 4.0 4.0 4.0 4.0 3.9 4.0  4.0 ^ 0.0  Experiment not analyzed (n=10) a On  a scale of 0 (healthy) to 4 (severe), rated 10 days after the FORL inoculation. Seedlings were grown on slants of 1% water agar in test tubes. They were treated with 0.1 ml of the potential elicitor, H 2 O or living nit B1 on day 5; the elicitor treatment continued over days 6,7,8 and 9. On day 3, all seedlings were inoculated with 1.6 x 10 3 microconidia of FORL (isolate 16).  87  Table 3.3. Mean disease severity ratings of plants treated with possible elicitors from nit B1-infested roots and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL) Disease severitya  Control^with FORL ^ ^ Root and stem washings ^ 0.0^2.2 Ground roots and stems ^0.0^2.7 2.5 0.0 Ground root and stem supernatant ^ 0.0^2.0 H2 O ^ 0.028^0.084 Ground roots and stems, not sterile Elicitor preparation  S.E. = 0.14^(n=4) eon a scale of 0 (healthy) to 4 (diseased), rated 2 weeks after  the FORL inoculation. Seedlings were grown on 1% water agar in petri plates. They were treated on day 4 with a drop of the potential elicitor, H 2 O or nit B1-infested, unsterilized ground roots and stems. A mycelial plug of FORL (isolate 16) was placed in the center of each plate (nothing placed in the center of control plates).  88  Table 3.4. Mean disease severity ratings of plants treated with the heat-released cell wall fraction of Fusarium strain nit B1 and then inoculated with Fusarium oxysporum f. sp. radicislycopersici (FORL) Cell wall preparation^ Disease severity a ^ Nit B1 cell wall pellet ^ 3.4 2.7 Nit B1 cell wall supernatant ^ FORL cell wall pellet 3.4 ^ 2.8 FORL cell wall supernatant ^ H2O 2.6 S.E. = 0.20 (n=4) L.S.D. = 0.533 (a=0.05) a an a scale of 0 (healthy) to 4 (diseased), rated 5 days after the FORL inoculation. Seedlings were grown on 1% water agar in Petri plates. On day 3 they were pricked at the crown and treated with one of the elicitor preparations. After 4 days they were inoculated with 1 x 10 4 microconidia of FORL (ssC 33).  89  DISCUSSION  The results from this chapter suggest that nit B1 must be alive in order to bring about biological control of crown and root rot in tomato. This is in accordance with several examples of biological control of Fusarium oxysporum pathogens by nonpathogenic Fusarium oxysporum strains in which induced resistance was demonstrated by other methods (Ogawa and Komada, 1981; Matta, 1989). Of course, it is possible that the elicitor activity was lost during preparation in these experiments. And it is possible that induced resistance is not a mechanism at all. However, models for plant-pathogen interactions are becoming more and more complex as more and more components of the interaction are discovered, and there are several possible explanations for the need for living nit B1. One explanation is that the elicitor of the host defense respose that limits fungal growth in the tomato root cortex comes from the plant cell wall rather than the fungal cell wall. Oligosaccharides that exhibit elicitor activity have been purified from soybean cell walls (Hahn et al., 1981; Davis et al., 1986). It is speculated that these are derived from the pectic polysaccharides present in all plant cell walls, and that they may be released by the partial degradation of these walls by the invading pathogen (Hahn et al., 1981). If this is the case in tomato as well, then nit B1 would have to be in the process of 90  infecting the root before any defense response was initiated. Microscopic studies of the partial colonization of the plant cortex by nonpathogenic Fusarium oxysporum or by wilt pathogens do show that invasion of the root by these fungi closely parallels the appearance of the structural barriers that limit their spread (Beckman and Talboys, 1981; Bishop and Cooper, 1983), suggesting that the induction of the defense reaction is related to the partial degradation of the plant cell walls. Different responses to nit B1 and FORL could be due to differences in the activity of these strains' cell wall-degrading enzymes. Another explanation for these results is that the response that is induced by the cell wall elicitor is a generalized response to the presence of any fungus. The B-1,3 glucan elicitor studied in other host-pathogen interactions is generally aspecific - it induces the same response in both susceptible and resistant host plants (Ayers et al., 1976). The real determinant of resistance, the factor that perhaps controls the timing of the response, may be located elsewhere, or perhaps, again, may be released only when the invading fungus is actually inside the host plant. If induced resistance is the mechanism of biological control in this system, nit B1 may stimulate the release of something else in the host, for example a signal molecule which, as mentioned in Chapter 1, "primes" the cells to respond more quickly to fungal invasion (Kuc, 1990; Van Peer and Schippers, 1991). 91  Although the results from this chapter do not provide evidence for induced resistance as a mechanism of biological control, they do not refute it either. Whatever the mechanism, it requires the presence of living nit B1 to be effective.  92  Chapter 4  Testing Nutrient Competition as a Possible Mechanism of Biological Control  INTRODUCTION  Both carbon and iron may play a role in the biological control of Fusarium oxysporum pathogens. The importance of carbon competition has not been proven, and is still a subject of debate (Paulitz, 1990). However, it has been suggested as a mechanism in some examples of biological control by nonpathogenic fusaria (Louvet et al., 1981). Carbon competition was tested in the present research by adding excess glucose to the growth medium. The glucose was added in combination with asparagine as a nitrogen source because previous research has shown that increasing the carbon concentration may lead to excessive fungal growth and depletion of nitrogen (Lockwood, 1988). There is much more evidence for iron competition, although iron competition is mainly a mechanism when the biological control agents are fluorescent pseudomonad bacteria, not nonpathogenic fusaria (Kloepper et al., 1980; Scher and Baker, 1982; Sneh et al., 1984). The form in which iron is available in the rhizosphere is as Fe 3+ . Most microorganisms produce siderophores which bind Fe ll' so that it can be taken up. These 93  siderophores have varying affinities for Fe 3+ , as represented by their stability constants (Scher and Baker, 1982). Certain bacteria produce siderophores with a higher stability constant than the siderophore, fusarine, produced by Fusarium species. There is a possibility that under certain conditions nonpathogenic Fusarium oxysporum strains may produce higher amounts of the siderophore, fusarine, than pathogenic strains (Lemanceau et al., 1986). Some of these strains may be more successful at taking up limited iron, and so may have a competitive advantage over the pathogen. Siderophores have receptors which allow them to be taken up by specific microorganisms (Bakker et al., 1990), so the siderophores produced by the nonpathogenic strains would not necessarily be available to the pathogen. Although there is no documented evidence of iron competition between strains of Fusarium oxysporum, there is evidence for enhanced biological control when nonpathogenic Fusarium oxysporum strains are combined with fluorescent pseudomonads (Park et al., 1988; Lemanceau and Alabouvette, 1991). This enhanced control may only be additive, with the fluorescent pseudomonads competing for iron and the nonpathogenic Fusarium strains bringing about control by another, unrelated mechanism. However, in order to survive and have a positive effect on biological control in the presence of the strongly iron-chelating pseudomonads, some nonpathogenic strains may have a more efficient mechanism for taking up iron than the pathogenic strains which are strongly suppressed in the presence 94  of the fluorescent pseudomonads. Competition for iron was tested in this research by adding excess iron to the growth medium in the form of FeEDTA. FeEDTA has a lower stability constant than fusarine (log 10 K = 25 as opposed to log 10 K = 29) (Scher and Baker, 1982), so Fe 3+ bound to EDTA should be available to the pathogen. In other experiments, EDDHA was added to the growth medium. The ligand EDDHA has a higher stability constant (log 10K = 33.9) than fusarine and, under conditions of extreme iron limitation, it should render iron unavailable to the pathogen and thus mimic the effects of the nonpathogen, if iron competition is indeed a major factor in the control.  95  METHODS  Competition for Carbon  In the first experiment, surface-sterilized tomato seeds were germinated on Whatman #3 filter paper that had been infested with 0.5 ml of a microconidial suspension of nonpathogenic  Fusarium strain nit B1 (prepared as described in Chapter 1). After 4 days the seedlings were transferred to plates of 1% water agar amended with various concentrations of glucose and asparagine in a 5:1 ratio (the media were prepared by making 1/2 dilutions from an initial solution containing 2.4 g 1 -1 glucose and 0.479 g 1 -1 L-asparagine, adding skim milk powder to raise the pH to 6.0, then adding the appropriate amount of agar). Each seedling was inoculated with a 50 Al drop of microconidial suspension of Fusarium oxysporum f. sp. radicis-lycopersici (FORL). In a second experiment, seedlings were germinated on sterile Whatman #3 filter paper wetted with 0.5 ml H 2 0. When they were 4 days old, they were transferred to glass microfibre (Whatman GF/A) wicks in test tubes containing 2 ml of Dutch fertilizer solution (Appendix B) amended with glucose and L-asparagine in a 5:1 ratio. The seedlings were inoculated with 0.1 ml of a nit B1 microconidial suspension. After 3 days they were inoculated with 0.1 ml of a FORL microconidial suspension. The seedlings were incubated in growth chambers under the conditions described in 96  Chapter 1.  Competition for Iron  1. Iron-chelator Stock Solutions Ferric chloride (FeCl 3 ) and the ferric-sodium salt of ethylenediaminetetraacetic acid (FeNaEDTA) were dissolved directly in H 2O. A solution of FeEDTA was made by dissolving equimolar amounts of FeCl 3 and EDTA in H 2 O, adjusting the pH to 7.0 with 2N KOH. Ethylenediaminine-O-hydroxyphenylacetic acid (EDDHA) was put in H 2 O and enough NaOH was added to make it dissolve. An equimolar amount of FeCl 3 was added to some of this to form FeEDDHA. Both EDDHA and FeEDDHA were adjusted to pH 6 using 1 N HC1 (Scher and Baker, 1982). All solutions were filter sterilized.  2. Biological Control in Excess Iron Solutions of FeNaEDTA or FeCl 3 were added in various amounts to Dutch fertilizer solution (pH 6) to give the desired iron concentrations. Seedlings (4 days old) were placed on glass microfibre wicks in test tubes containing 2 ml of the ironfertilizer solution, and inoculated with 0.1 ml of a suspension of nit B1 microconidia. After 3 days they were inoculated with 0.1 ml of a suspension of FORL microconidia, along with an extra 0.2 ml of the iron-fertilizer solution. Plants were treated again with one of the iron solutions after 6 days. 97  In a second experiment a solution of FeEDTA was added to molten agar in test tubes to give various iron concentrations, before the tubes were slanted. Seedlings (4 days old) were placed on the agar slants and inoculated with 0.1 ml of a suspension of nit B1 microconidia or H 2 O. After 3 days they were inoculated with 0.2 ml of a suspension of FORL microconidia or H 2 O. Some of the plants were weighed. All the plants that had been treated with both nit B1 and FORL were taken off their slants, their roots were cut off and rinsed in 2 ml sterile H 2 O, and 1 ml of the rinse water was plated onto each type of medium (FMM, selective for FORL, and FMM-C, selective for nit B1).  3. Disease Symptoms in the Presence of a Strong Iron Chelator Solutions of EDDHA, FeEDDHA, FeC1 3 or a weak solution of NaC1 (made from the same amount of NaOH and HC1 as was added to neutralize the iron chelator solutions) were added to Dutch fertilizer solution (pH 6) to give the desired chelator concentrations. Seedlings (4 days old) were placed on glass microfibre wicks in test tubes containing 0.2 ml of the iron or iron - chelator solution. The following day, all seedlings were inoculated with 0.1 ml of a suspension of FORL microconidia. The second experiment was the same except that the solutions were added to sterile distilled water instead of Dutch fertilizer solution. The plants were inoculated with 0.1 ml of a suspension of FORL microconidia.  98  RESULTS  Competition for Carbon  Biological control was not affected when plants were grown with excess glucose and nitrogen in the medium (Table 4.1, and ANOVA, Table A14). In both experiments the nit B1 x FORL interaction was highly significant, while the nit B1 x FORL x glucose level was not. In the first experiment nit B1 itself caused some disease symptoms at the higher glucose concentrations, as indicated by a significant nit B1 x glucose interaction. These symptoms were still less severe than those caused by FORL alone. In the second experiment nit B1 did not cause significant disease symptoms at any glucose level.  Competition for Iron  1. Biological Control in Excess Iron Biological control was not affected when plants were grown in a medium containing excess iron, in the form of FeEDTA or FeC1 3 (Table 4.2 and ANOVA, Table A15). In the first experiment, high levels of NaFeEDTA (500 Ag m1 -1 ) caused increased disease severity ratings, whether or not FORL was present. FeC1 3 did not stay in solution, but despite the iron precipitate FeC1 3 levels had no effect on disease severity ratings (a=0.05). 99  Table 4.1. Biological control of Fusarium crown and root rot by nonpathogenic Fusarium strain nit B1 in the presence of excess glucose and nitrogen Disease severity a Glucose concentration (mg m1 -1 )  Control  With nit B1  with FORL  no FORL  with FORL  0.1 0.1 0.0 0.0 0.0  4.0 4.0 3.9 4.0 3.9  2.1 0.6 0.3 0.2 0.1  2.4 0.7 0.1 0.1 0.1  0.2 0.4 0.4 0.3 0.4  4.0 4.0 4.0 4.0 4.0  0.0 0.5 0.2 0.2 0.1  0.2 0.0 0.3 0.1 0.0  no FORL b  Experiment 1 2.4 1.2 0.6 0.3 0.0 S.E.^= 0.16^(n=4) Experiment 2 2.4 1.2 0.6 0.3 0.0 S.E. = 0.11 (n=10) a sOn  a scale of 0 (healthy) to 4 (severe), rated 10 days after the FORL inoculation b Fusarium oysporum f. sp. radicis-lycopersici Experiment 1: Seedlings were grown on plates of 1% water agar amended with glucose and nitrogen in a 1:5 ratio. They were inoculated at seeding with H 2 O or 3 x 10 2 conidia of nit Bl, and 4 days later with NO or 2 x 10 2 conidia of FORL (isolate 16). Experiment 2: Seedlings were grown on glass microfibre wicks in Dutch fertilizer solution. They were inoculated on day 4 with H 2 O or 7 x 10 3 conidia of nit Bl, and 3 days later with H 2 O or 1 x 10 2 conidia of FORL (isolate 16).  100  Table 4.2. Biological control of Fusarium crown and root rot by nonpathogenic Fusarium strain nit B1 in the presence of excess iron Disease Severity a Control^ With nit B1 Iron source^no FORL b  FORL  no FORL  FORL  Experiment 1 NaFeEDTA (gg m1 -1 ) 500 100 50 FeC1 3^(gg m1 -1 ) 78 31 16  1.8 0.4 0.2  1.4 0.2 0.4  0.8 0.2 0.2  0.0 0.2 0.2  H2O  0.0  0.2  S.E.^= 0.23^(n=5) L.S.D.^= 0.91^(a=0.05) Experiment 2 FeEDTA (Ag m1 1 ) 1000 500 100 50  0.1 0.1 0.0 0.0  4.0 4.0 4.0 4.0  0.3 0.0 0.1 0.0  0.3 0.0 0.0 0.0  H2 O  0.0  4.0  0.0  0.0  -  S.E.^= 0.054^(n=10) L.S.D. = 0.077 (a=0.05) a On  a scale of 0 (healthy) to 4 (severe), rated 10 days after the FORL inoculation b Fusarium oxysporum f. sp. radicis-lycopersici Experiment 1: Seedlings were grown on glass microfibre wicks in Dutch fertilizer solution. They were inoculated at day 4 with H 2 O or 1 x 10 3 conidia of nit B1, and at day 7 with H 2 O or 2 x 10 3 conidia of FORL (isolate 16). Experiment 2: Seedlings were grown on slants of 1% water agar amended with FeEDTA; they were inoculated as in Experiment 1.  101  In the second experiment high levels of FeEDTA caused a slight increase in crown and root rot - like disease severity ratings (some plants in treatments receiving 1000 Ag m1 -1 had small browning patches on the crown that were rated "1"), but these, again, were independent of the FORL treatment. In this experiment, all the plants that had been treated with 500 or 1000  pg m1-1 of FeEDTA were very sickly. Their seed coats had not come off, the leaves were stuck together, and the roots were lightbrown and mushy. These symptoms appeared both in control plants and in fungus-inoculated plants and in fact appeared to be more severe in the controls. The symptoms looked quite different from the symptoms of crown and root rot which is why they were not considered in the ratings. Only plants that had dark brown lesions were given a positive rating. Plants grown in 1,000 pg m1 -1 FeEDTA weighed significantly less than those grown in 50 pg m1 -1 FeEDTA (Table 4.3 and ANOVA, Table A16). This weight loss masked the biological control by nit Bl. At the lower FeEDTA level biological control was indicated by increased root weights in the protected plants. The degree of surface colonization of five protected plants from each of the iron treatments was measured (Table 4.4 and ANOVA, Table A17). Although significantly more (a=0.01) CFU's of nit B1 than of FORL were isolated, there was no difference in CFU's isolated from plants grown at different iron levels.  102  ^ ^  Table 4.3. Mean root weights (mg) of tomato seedlings grown in iron-rich media, inoculated with nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) FeEDTA 50 Ag m1 -1^FeEDTA 1000 Ag m1 -1 no FORL^with FORL no FORL^with FORL ^ ^ no nit B1^59.8^ 15.6^17.8^9.53 with nit B1^54.0 12.6^14.4 52.4 S.E. = 2.2 (n=10) L.S.D. = 6.17 (a=0.05) Seedlings were grown on slants of 1% water agar amended with FeEDTA. They were inoculated at day 4 with H 2 O or 1 x 10 3 conidia of nit Bl, and at day 7 with H 2 O or 2 x 10 3 conidia of FORL.  Table 4.4. Colony forming units of nonpathogenic Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici washed from the roots of tomato seedlings grown in iron-rich medium FeEDTA concentration (Ag m1 -1 ) 1000^ 500^ 100^ 50^  CFU's in 1 ml rinse water Nit B1^FORL 275^ 174 326^ 65 286^ 89 234^ 46  S.E. = 45 (n=5) L.S.D. = 154 (a=0.05) Seedlings were grown on slants of 1% water agar amended with FeEDTA. They were inoculated at day 4 with H 2 O or 1 x 10 3 conidia of nit Bl, and at day 7 with H 2 O or 2 x 10 3 conidia of FORL. Colony forming units were recovered 10 days after the FORL inoculation.  103  2.Disease Severity in the Presence of a Strong Iron Chelator When EDDHA was added to the growing medium at a concentration of 400 Ag m1 -1 , there was a slight, but significant decrease in disease severity ratings of FORL-inoculated plants (Table 4.5 and ANOVA, Table A18). This high concentration of EDDHA had a deleterious effect on the plants, however. They developed a slight yellowing which was rated "1".  104  Table 4.5. Mean disease severity ratings of tomato seedlings treated with the iron chelator EDDHA and inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL)  Disease severity a Iron chelator  no FORL^with FORL  EDDHA (Ag m1 -1 ) 400 200 100 50  1.0 0.8 0.6 0.2  3.2 3.8 3.8 4.0  FeEDDHA (Ag m1 -1 ) 400 200 100 50  0.0 0.2 0.4 0.6  4.0 4.0 4.0 4.0  (H2O) (NaC1)  0.6 0.4  4.0 4.0  S.E.^= 0.23^(n=5) L.S.D.^= 0.63^(a=0.05) a On  a scale of 0 (healthy) to 4 ( severe), rated 8 days after the FORL inoculation. Seedlings were grown on glass microfibre wicks in iron chelator solutions. They were inoculated with H 2 O or 8 x 10 3 conidia of FORL (SSc 33).  105  DISCUSSION  Competition for Carbon  Carbon competition was not demonstrated in this section. Adding glucose to water agar or nutrient solution at concentrations of up to 2.4 mg m1 -1 (with 0.48 mg m1 -1 asparagine) had no effect on the observed biological control (Table 3.1). This is in contrast with the results from experiments dealing with carbon competition in soils. Sneh et al. (1984) found that adding glucose to a Fusarium wilt-suppressive soil led to an increase in the percent germination of F. oxysporum f. sp.  cucumerinum chlamydospores: this reached a maximum (which was still only about half the percent germination of these chlamydospores in wilt conducive soils) when the glucose concentration was approximately 0.8 mg g -1 soil, with 0.2 mg asparagine g -1 soil. Sivan and Chet (1989) found that adding glucose to a soil in which Trichoderma harzianum was the biological control agent led to an increase in the percent germination of F. oxvsporum f. sp. melonis chlamydospores: this reached a maximum, equal to the percent germination of these chlamydospores in control soils, when the glucose concentration was approximately 0.3 mg g -1 soil, with 0.06 mg asparagine g -1 soil. The results from the carbon competition experiment in this research are not surprising when one considers the conditions 106  under which the biological control occured. The pathogen inoculum was in the form of microconidia, which have lower nutrient requirements for germination than chlamydospores (Griffin, 1981) and the only microorganisms present were the two strains of  Fusarium oxysporum so it is highly unlikely that carbon was limited. In raw soils carbon competition is more probable: the combined activities of all the microorganisms in the soil might limit carbon and so enhance the competition between the biological control agent and the pathogen (Alabouvette, 1990). High amounts of glucose in the growing medium led to some pathogenicity by nit Bl. Researchers studying the defense responses of various plants have sometimes found that these responses are less effective at excluding the invading microorganisms when the plants are grown on nutrient-rich media (Beckman and Talboys, 1981). This may be due to increased virulence of the microorganisms or to a breakdown in the host defense system. As mentioned before, it is possible that nit B1 does invade the cortical cells of the tomato root to a limited extent. It is possible, then, that the partial breakdown of the defenses that normally prevent this microorganism from destroying the cortex led to the slight increase in disease severity ratings, without leading to a complete breakdown in defenses and invasion by either nit B1 or FORL. When this experiment was repeated, the plants were grown on filter wicks in test tubes rather than on agar in petri plates, so that the whole root was not constantly sitting on the glucose medium. This may explain 107  why the glucose-induced symptoms did not appear in this second experiment.  Competition for Iron  There was no evidence for iron competition between nit B1 and FORL. Adding excess iron, in the form of FeEDTA at up to 1000 Ag m1 -1 (5.9 AM Fe m1 -1 ) or FeCl 3 at up to 78 Ag m1 -1 (0.4 AM Fe m1 -1 ) had no effect on biological control or on the colonization of tomato roots by either nit B1 or FORL. This is in contrast with the results of Kloepper et al. (1980), who found that Fusarium-suppressive soil was made conducive by the addition of 0.5 AM m1 -1 FeEDTA and Scher and Baker (1982) who found that FeEDTA at 100 Ag g -1 soil nullified the biological control of F. oxysporum f. sp. lini and F. oxysporum f. sp. conglutinans by Pseudomonas Dutida. As with carbon competition, iron competition is more likely to be a factor in raw soils in which the pathogen inoculum is in the form of chlamydospores than in the in vitro conditions under which biological control was demonstrated here. Iron competition is also more probable when there is a qualitative difference in the stability constants of the siderophores produced by the biological control agents and the pathogen (as is the case when the biological control agent is a fluorescent pseudomonad) than when the possible difference in siderophore production is only quantitative. Furthermore, the Fusarium siderophore, fusarine, is produced only under conditions 108  of extreme iron limitation, less than 10  -1  mg L -1 (Lemanceau et  al., 1986). The iron content of the fertilizer solutions used in this research was on the order of 1 mg L -1 . The importance of differing degrees of fusarine production among different Fusarium  oxysporum strains in biological control even in suppressive soils is questionable (Lemanceau et al., 1986). Adding a very high concentration (400 Ag m1 -1 ) of the strong Fe 3+ chelator, EDDHA, did decrease the disease severity ratings of FORL-inoculated plants somewhat (Table 3.5). This chelator binds Fe ll' at a higher stability constant than fusarine (log m K = 33.9 as opposed to log 10 K = 29) yet its effect, even at this high concentration, was much less pronounced than the effect of nit Bl. This is again in contrast with the results found in experiments in which iron competition is a putative mechanism. For example, Scher and Baker (1984) found that the incidence of wilt caused by F. oxysporum f. sp. lini, F. oxysporum f. sp.  conglutinans and F. oxysporum f. sp. cucumerinum was reduced to about half the incidence in the control when EDDHA was added at a concentration of 100 Ag g -1 soil. The fact the EDDHA had only a very slight effect on disease severity ratings gives more support to the idea that iron was not at all limited in the experiments done here, and makes it even more unlikely that the small amounts of fusarine that might be produced under the conditions of these experiments contributed to the observed biological control. However, many complex factors influence iron availability and siderophore production in soil 109  and growth media, (Scher and Baker, 1984; Lemanceau et al., 1986; Simeoni et al., 1987). More precise manipulation of iron levels would be neeeded to determine the exact role of fusarine production in intraspecific competition among different strains of Fusarium oxysporum.  110  GENERAL DISCUSSION  In the experimental system developed for this research, tomato seedlings were grown on plates of water agar or on wicks in test tubes containing fertilizer solution. This system was convenient to use and allowed extensive characterization of the biologial control of tomato crown and root rot by a nonpathogenic strain of Fusarium oxysporum. Because a nit mutant of the nonpathogenic strain (nit B1) was used, the colonization of tomato roots by this strain and the pathogen, Fusarium oxysporum f. sp. radicis-lycopersici (FORL) could be observed. The results of the experiments were discussed in relation to four commonly suggested mechanisms of biological control: antibiosis, nutrient competition, induced resistance and competition for infection sites. When nit B1 was inoculated to seedling roots at least 3 days before FORL, there was good disease control even when the initial inoculation concentration of nit B1 was 60 times smaller than the initial inoculum concentration of FORL. This was characteristic of a mechanism in which nit B1 was inducing a delayed defense response in the whole root, making it resistant to infection by FORL. As the lag between inoculations decreased, higher concentrations of nit B1 were needed to bring about control. When the two fungi were inoculated to the roots simultaneously the inoculum concentration of nit B1 had to be 10 times the concentration of FORL for there to be any reduction of disease 111  symptoms. In this latter case, it was hypothesized that nit B1 was competing with FORL for occupation of a limited number of colonization sites on the root, or else that the resistance reaction was limited to a few cells surrounding the nit B1 propagule, so that the extent of this reaction would depend on the inoculum density of these propagules. When nit B1 was inoculated to tomato seedling roots 4 days before FORL, the rate of increase of nit B1 on the roots was greater than the rate of increase of FORL. This indicated that competition for infection sites within the root might be a factor in the biological control. However, despite its greater rate of increase, nit B1 did not prevent FORL from colonizing the root, and in preliminary experiments good biological control was observed even when equal numbers of FORL and nit B1 colony forming units were recovered from the roots. As well, there was no indication that colonization sites on the root were limited. Again, it seemed that nit B1 was eliciting a response in the roots that made them resistant to cortical rot. Direct or host-mediated antibiosis was not indicated, since nit B1 and FORL increased at similar rates on the root surface. When sterile elicitor preparations made from filtrates of nit B1 cultures were tested for their biological control abilities, the results were all negative. It was suggested that nit B1 had to actually colonize the root to a limited extent before it could bring about control. Adding excess glucose and iron to the medium did not nullify 112  the control, so competition for these nutrients was not supported. A hypothetical model for the biological control observed in this system would be that the nonpathogen, nit Bl, colonizes the root to a limited extent and, by its presence, triggers a response in the roots that makes them resistant to subsequent infection by FORL. If this is the mechanism then some interesting questions remain to be answered: to what degree does nit B1 colonize tomato roots - does it colonize the whole cortex or just the epidermis and hypodermis? Is the resistance response limited to cells directly adjacent to nit B1 hyphae or is the whole root system affected? Do the same defense responses that prevent nit B1 from spreading in the whole cortex (for example, mechanical barriers and cell wall thickening) appear in all root cells or are these cells simply "primed" to respond quickly to subsequent infection? Induced resistance has been demonstrated in several systems involving the biological control of Fusarium oxysporum pathogens by avirulent strains of the pathogen or by alien pathogenic  Fusarium oxysporum strains (Biles and Martyn, 1989; Kroon et al., 1990). However, this mechanism is seldom suggested for the biological control that is brought about by nonpathogenic  Fusarium oxysporum strains in suppressive soils. Some authors claim that induced resistance is an illogical mechanism for biological control in suppressive soils since the soils are suppressive to wilts in plants from many different botanical host 113  families, yet induced resistance usually depends on a specific host/parasite interaction (Paulitz, 1990). As well, the characteristics of biological control by avirulent strains of the pathogen or by alien pathogenic strains are different from those of biological control by nonpathogenic strains from suppressive soils. In the former, the protection is usually transitory while in the latter the protection usually lasts the whole growing season (Louvet et al., 1981). It would appear, then, that there are two general systems of biological control of Fusarium wilt pathogens by strains of Fusarium oxysporum, one, such as the system studied in this research, in which induced resistance is the primary mechanism, and one, such as that found in suppresive soils, in which a form of competition may be the most important factor in the control. But it is possible that the same mechanism is responsible for control in both systems. The different characteristics of the two systems can be explained when inoculum densities and inoculum methods in the two types of experiments are compared. In "induced resistance" experiments the inoculum density of the inducing organism is usually less than or equal to the inoculum density of the pathogen (ie. Wymore and Baker, 1982; Biles and Martyn, 1989; Jorge et al., 1992) whereas in experiments dealng with suppressive soil the nonpathogen is usually added at an inoculum density 10 times that of the pathogen (ie. Paulitz et al., 1987; Mandeel and Baker, 1991; Lemanceau et al., 1992). Also, in induced resistance experiments the plants are inoculated by dipping them in a suspension of 114  microconidia of the inducing strain, then replanting them in sterile soil, then uprooting them again and redipping them in a suspension of pathogen microconidia (ie. Biles and Martyn, 1989; Kroon et al., 1991; Jorge et al., 1992); in experiments dealing with suppressive soil the nonpathogenic strains and the pathogenic strains are mixed into the soil before the seeds are planted (ie. Paulitz et al., 1987; Lemanceau et al., 1992; Mandeel and Baker, 1991). Thus in the latter, the growing root is constantly encountering a new source of nonpathogenic inoculum and the ratio of this to the pathogen does not change, whereas in the former, depending on the length of delay between inoculations and the ability of the inducing organism to move downwards on the root, the new growing tips may be more or less free of the inducing organsim. If biological control involves a short-term induction of a resistance reaction that depends on the constant presence of the inducing organism on the root for this reaction to be persistent, then only inoculum procedures such as those used in suppressive soil experiments will allow this protection to be long-term. The logical objection to induced resistance as a mechanism of control in suppressive soils is overcome when one considers the fact that many nonpathogenic, alien pathogenic and avirulent strains of Fusarium oxysporum can colonize the cortex of a variety of host plants, without causing any kind of root rot or wilt (Katan, 1971; Gordon et al., 1989; Farias and Griffin, 1990). It is not illogical to suggest that roots have a 115  generalized reaction to Fusarium oxysporum and that normally this resistance response limits Fusarium oxysporum strains to a few cells or to the intercellular spaces in the cortex. Certain strains have developed the ability to overcome this reaction in certain hosts, either in the cortex (in which case they are root rot pathogens) or in the stele (in which case they are wilt pathogens). However, any nonpathogenic or alien pathogenic  Fusarium oxysporum strain that is an efficient root colonizer, both on the surface and in the cortex, has the potential to provide some protection against any of these Fusarium oxysporum pathogens.  116  BIBLIOGRAPHY Alabouvette, C. 1990. Biological control of Fusarium wilt pathogens in suppressive soils. In: Hornby, D. (Ed.) Biological Control of Soil-Borne Plant Pathogens. CAB International, Wallingford, U.K. pp. 27-43. Alabouvette, C. 1991. Suppressive soils and practical application of biological control of Fusarium diseases. In: Bay-Peterson, J. (Ed.) The Biological Control of Plant Diseases, FFTC Book Series No. 42. pp. 120-129. Alabouvette, C., Rouxel, F. and Louvet, J. 1979. Characteristics of Fusarium wilt-suppressive soils and prospects for their utilization in biological control. In: Schippers, B. and Gams, W. (Eds.) Soil-Borne Plant Pathogens. Academic Press, New York. pp 165-182. Ayers, A., Ebel, J., Valent, B. and Albershein, P. 1976. Fractionation and biological activity of an elicitor isolated from the mycelial walls of Phytopthora megasperma var. soiae. Plant Physiol. 57: 760-765. Baker, R. 1968. Mechanisms of biological control of soil-borne pathogens. Ann. Rev. Phytopathol. 6: 263-294. Baker, R. 1990. An overview of current and future strategies and models for biological control. In: Hornby, D. (Ed.) Biological Control of Soil-Borne Plant Pathogens. CAB International, Wallingford, U.K. pp. 375-388. Bakker, P., Van Peer, R., and Schippers, B. 1990. Specificity of siderophores and siderophore receptors and biocontrol by Pseudomonas species. In: Hornby, D. (Ed.) Biological Control of Soil-Borne Plant Pathogens. CAB International, Wallingford, U.K. pp. 131-142. Banihashemi, Z. and deZeeuw, D.J. 1975. The behaviour of Fusarium oxysporum f. sp. melonis in the presence and absence of host plants. Phytopathology 65: 1212-1217. Beckman, C.H. and Talboys, P.W. 1981. Anatomy of resistance. In: Mace, M.E., Bell, A.A. and Beckman, C.H. (Eds.) Fungal Wilt Diseases of Plants. Academic Press, New York, pp. 487-521. Belhadj, R., Charest, P.M. and Theriault, G. 1987. Electrophoretic comparison of the protein composition of two closely related fungi: Fusarium oxysporum f. sp. radicis lycopersici (FORL) and F. oxysporum f. sp. lycopersici (FOL). Can. J. Plant Pathol. 9: 272-273. (Abstr.). 117  Benhamou, N., Grenier, J., Asselin, A. and Legrand, M. 1989. Immunogold localization of B-1,3-glucanases in two plants infected by vascular wilt fungi. Plant Cell 1: 1209-1221. Benhamou, N., Joosten, M. and De Wit, P. 1990. Subcellular localization of chitinase and of its potential substrate in tomato root tissues infected by Fusarium oxysporum f. sp. radicis-lycopersici. Plant Physiol. 92: 1108-1120. Benhamou, N., Mazau, D. and Esquerre-Tugayê, M.-T. 1990. Immunocytochemical localization of hydroxyproline-rich glycoproyeins in tomato root cells infected by Fusarium oxysoporum f. sp. radicis-lycopersici: study of a compatible interaction. Phytopathology 80: 163-173. Biles, C.L. and Martyn, R.D. 1989. Local and systemic resistance induced in watermelons by formae speciales of Fusarium oxysporum. Phytopathology 79: 856-860. Bishop, C.D. and Cooper, R.M. 1983. An ultrastructural study of root invasion in three vascular wilt diseases. Physiol. Plant Pathol. 22: 15-27. Booth, C. 1971. The Genus Fusarium. Commonwealth Mycological Institute, Kew., U.K. 237 pp. Bowen, G.D. and Rovira, A.D. 1976. Microbial colonization of plant roots. Ann. Rev. Phytopathol. 14: 121-144. Boyer, A. and Charest, P.M. 1989. Use of lectins for differentiating between Fusarium oxysporum f. sp. radicislycopersici and F. oxysporum f. sp. lycopersici in pure culture. Can. J. Plant Pathol. 11: 14-21. Brammal, R.A. and Higgins, V.J. 1988. A histological comparison of fungal colonization in tomato seedlings susceptible or resistant to Fusarium crown and root rot disease. Can. J. Bot. 66: 915-925. Brammal, R.A. and McKeown, A.W. 1989. An occurrence in Ontario of Fusarium crown and root rot disease in field-grown processing tomatoes originating from multicelled tray transplants. Can. J. Plant Pathol. 11: 76 (Abstr.). British Columbia Ministry of Agriculture and Fisheries. 1987. Greenhouse vegetable production guide for commercial growers. Province of British Columbia. 38 pp. Burgess, L.W. 1981. General ecology of the fusaria. In: Nelson, P.E., Tousson, T.A. and Cook, R.J. (Eds.). Fusarium: Diseases, Biology and Taxonomy. Penn. State Univ. Press, University Park. pp. 225-235. 118  Caron, M., Fortin, J.A. and Richard, C. 1986. Effect of phosphorus concentration and Glomus intraradices on Fusarium crown and root rot of tomatoes. Phytopathology 76: 942-946. Charest, P.M., Ouellette, G.B. and Pauze, F.J. 1984. Cytological observations of early infection process by Fusarium oxysporum f. sp. radicis-lycopersici in tomato plants. Can. J. Bot. 62: 12321244. ChOrif, M. and Benhamou, N. 1990. Cytochemical aspects of chitin breakdown during the parasitic action of a Trichoderma sp. on Fusarium oxysporum f. sp. radicis-lycopersici. Phytopathology 80: 1406-1414. Copeman, R.J., Eaton, G.W. and Mauza, B.E. 1990. Cross protection increases yields from Fusarium crown and root rot infected tomato plants grown in sawdust culture. In: McClay, A.S. (Ed.) Proceedings of the Workshop on Biological Control of Pests in Canada. Alberta Environmental Center, Vegreville, Alberta. p. 95. Correll, J.C., Kuttich, C.J.R. and Leslie, J.F. 1987. Nitrate non-utilizing mutants of Fusarium oxysporum and their use in vegetative compatibility tests. Phytopathology 77: 1640-1646. Cove, D.J. 1976. Chlorate toxicity in Aspergillis nidulans: the selection and characterization of chlorate-resistent mutants. Heredity 36: 191-203. Davies, K., Darvill, A., Albershein, P. and Dell, A. 1986. Oligogalacuronides released from sodium polypectate by endopolygalacturonic lyase are elicitors of phytoalexins in soybean. Plant Physiol. 80: 568-577. Elad, Y. and Baker, R. 1985. The role of competition for iron and carbon in suppression of chlamydospore germination of Fusarium spp. by Pseudomonas spp. Phytopathology 75: 1053-1059. Elmer, W.H. and Lacy, M.L. 1987. Effects of crop residues and colonization of plant tissues on propagule survival and soil populations of Fusarium oxysporum f. sp. apii race 2. Phytopathology 77: 381-387. Farias, G.M. and Griffin, G.J. 1990. Extent and pattern of early soybean seedling colonization by Fusarium oxysporum and F. solani in naturally infested soil. Plant and Soil 123: 59-65. Foster, R.C., Rovira, A.D. and Cook, T.W. 1983. Ultrastructure of the root soil interface. Amer. Phytopathological Society. St. Paul, Minn. 157 pp.  119  Garrett, S.D. 1979. The soil-root interface in relation to disease. In: Harley, J.L. and Russell, R.S. (Eds.) The Soil-Root Interface. Academic Press, London. pp. 301-303. Gordon, T.R., Okamoto, D. and Jacobson, D.J. 1989. Colonization of muskmelon and non-susceptible crops by Fusarium oxysporum f. sp. melonis and other species of Fusarium. Phytopathology 79: 1095-1100. Griffin, G.J. 1981. Physiology of conidium and chlamydospore germination in Fusarium. In: Nelson, P.E., Tousson, T.A. and Cook, R.J. (Eds.). Fusarium: Diseases, Biology and Taxonomy. Penn. State Univ. Press, University Park. pp. 331-339. Hadar, E., Katan, J. and Katan, T. 1989. The use of nitrate nonutilizing mutants and a selective medium for studies of pathogenic strains of Fusarium oxysporum. Plant Disease 73: 800803. Hahn, M. Darvill, A. and Albershein, P. 1981. The endogenous elicitor, a fragment of a plant cell wall polysaccharide that elicits phytoalexin accumulation in soybeans. Plant Physiol. 68: 1161-1169. Hartman, J.R. and Fletcher, J.T. 1991. Fusarium crown and root rot of tomatoes in the U.K. Plant Pathology 40: 85-92. Hoagland, D.R. and Arnon, D.I. 1938. The water culture method for growing plants without soil. U. of California Ag. Expt. Station Circular No. 347. pp. 1-39. Jarvis, W.R. 1988. Fusarium crown and root rot of tomatoes. Phytoprotection 69: 49-64. Jarvis, W.R. 1989. Epidemiology of Fusarium oxysporum f. sp. radicis-lycopersici. In: Tjamos, E.C. and Beckman, C. (Eds.) Vascular Wilt Diseases of Plants. NATO ASI Series Vol. H28. Springer-Verlag, Berlin. pp. 397-412. Jarvis, W.R. and Shoemaker, R.A. 1978. Taxonomic status of Fusarium oxysporum causing foot and root rot of tomato. Phytopathology 68: 1697-1680. Jarvis, W.R. and Thorpe, H.J. 1976. Susceptibility of lycopersici spp. and hybrids to Fusarium crown and root rot. Plant Dis. Reptr. 60: 1027-1031. Jorge, P.E., Green, R.J. and Chaney, W.R. 1992. Inoculation with Fusarium and Verticillium to increase resistance in Fusariumresistant tomato. Plant Dis. 76: 340-343.  120  Kasenberg, T.R. and Traquair, J.A. 1988. Effects of phenolics on growth of Fusarium oxysporum f. sp. radicis-lycopersici in vitro. Can. J. Bot. 66: 1174-1177. Katan, J. 1971. Symptomless carriers of the tomato Fusarium pathogen. Phytopathology 61: 1213-1217. Kloepper, J.W., Leong, M. and Schroth, M.N. 1980. Pseudomonas siderophores: a mechanism explaining disease-suppressive soils. Curr. Microbiol. 4: 317-320. Komada, H. 1990. Biological control of Fusarium wilts in Japan. In: Hornby, D. (Ed.) Biological Control of Soil-Borne Plant Pathogens. CAB International. Wallingford, U.K. pp. 65-75. Kroon, B.A.M., Scheffer, R.J. and Elgersma, D.M. 1991. Induced resistance in tomato plants against Fusarium wilt invoked by Fusarium oxysporum f. sp. dianthi. Neth. J. P1. Path. 97: 401408. Kuc, J. 1990. Immunization for the control of plant disease. In: Hornby, D. (Ed.) Biological Control of Soil-Borne Plant Pathogens. CAB International, Wallingford, U.K. pp. 355-373. Leary, J.V. and Endo, R.M. 1971. A Fusarium-induced root rot of staked tomatoes. Phytopathology 61: 900 (Abstr.). Lemanceau, P. and Alabouvette, C. 1991. Biological control of Fusarium diseases by fluorescent Pseudomonas and nonpathogenic Fusarium. Crop Protection 10: 279-286. Lemanceau, P., Alabouvette, C. and Meyer, J.M. 1986. Production of fusarine and iron assimilation by pathogenic and nonpathogenic Fusarium. In: Swinburne, T.R. (Ed.) Iron, Siderophores and Plant Diseases. NATO ASI Series, Series A. Plenum Press: New York. pp. 251-259. Lemanceau, P., Bakker, P., Jan de Kogel, W. Alabouvette, C. and Schippers, B. 1992. Effect of pseudobactin 358 production by Pseudomonas putida WCS358 on suppression of Fusarium wilt of carnations by nonpathogenic Fusarium oxysporum Fo47. Appl. and Environ. Microbiol. 58: 2978-2982. Lockwood, J.L. 1988. Evolution of concepts associated with soilborne plant pathogens. Ann. Rev. Phytopathol. 26: 93-121. Louter, J.H. 1986. Cross-Protection of Greenhouse Tomatoes against Fusarium Crown and Root Rot. Master's Thesis, University of Guelph. 117 pp.  121  Louter, J.H. and Edgington, L.V. 1990. Indications of crossprotection against Fusarium crown and root rot of tomato. Can. J. Plant Pathol. 12: 283-288. Louvet, J., Alabouvette, C. and Rouxel, F. 1981. Microbiological suppressiveness of some soils to Fusarium wilts. In: Nelson, P.E., Tousson, T.A. and Cook, R.J. (Eds.). Fusarium: Diseases, Biology and Taxonomy. Penn. State Univ. Press, University Park. pp. 261-275. Madhosingh, C. and Starratt, A.N. 1987. Comparison of antibiotic tolerance, lipids and respiration in the tomato pathogens Fusarium oxysporum lycopersici and Fusarium oxysporum radicislycopersici. J. Environ. Sci. Health B22: 731-749. Mandeel, Q. and Baker, R. 1991. Mechanisms involved in biological control of Fusarium wilt of cucumber with strains of nonpathogenic Fusarium oxysporum. Phytopathology 81: 462-469. Marois, J.J., Mitchell, D.J. and Sonoda, R.M. 1981. Biological control of Fusarium crown rot of tomato under field conditions. Phytopathology 71: 1257-1260. Mas, P., Molot, P.M. and Risser, G. 1981. Fusarium wilt of muskmelon. In: Nelson, P.E., Tonssou, T.A. and Cook, R.J. (Eds.) Fusarium: Diseases, Biology, and Taxonomy. Penn. State Univ. Press, University Park. pp. 169-177. Matta, A. 1989. Induced resistance to Fusarium wilt diseases. In: Tjamos, E.C. and Beckman, C. (Eds.) Vascular Wilt Diseases of Plants. NATO ASI Series 28: 175-196. Menzies, J.G., Koch, C. and Seyward, F. 1990. Additions to the host range of Fusarium oxysporum f. sp. radicis-lycopersici. Plant Dis. 74: 569-572. Mihuta-Grimm, L., Erb, W.A. and Rowe, R.C. 1990. Fusarium crown and root rot of tomato in greenhouse rock wool systems: sources of inoculum and disease management with benomyl. Plant Dis. 74: 996-1002. Nutter, F.W., Warren, C.G., Wells, O.S. and MacHardy, W.E. 1978. Fusarium foot and root rot in New Hampshire. Plant Dis. Reptr. 62: 976-978. Ogawa, K. and Komada, H. 1986. Induction of systemic resistance against Fusarium wilt of sweet potato by nonpathogenic Fusarium oxysporum. Ann. Phytopath. Soc. Japan 52: 15-21.  122  Park, C.S., Paulitz, T.C. and Baker, R. 1988. Biocontrol of Fusarium wilt of cucumber resulting from interactions between Pseudomonas Dutida and nonpathogenic isolates of Fusarium oxysporum. Phytopathology 78: 194-199. Parkinson, D. 1967. Soil microorganisms and plant roots. In: Burges, A. and Raw, F. (Eds.) Soil Biology. Academic Press, New York. pp. 449-478. Paulitz, T.C. 1990. Biochemical and ecological aspects of competition in biological control. In: Baker, R. and Dunn, P.E. (Eds.) New Directions in Biological Control: Alternatives for Suppressing Agricultural Pests and Diseases. Alan R. Liss, New York. pp. 713-724. Paulitz, T.C., Park, C.S. and Baker, R. 1987. Biological control of Fusarium wilt of cucumber with nonpathogenic isolates of Fusarium oxysporum. Can. J. Microbiol. 33: 349-353. Roby, D., Toppan, A. and Esquerre-Tugaye, M.-T. 1985. Elicitors of fungal and of plant origin trigger the synthesis of ethylene and of cell wall hydroxyproline-rich glycoprotein in plants. Plant Physiol. 77: 700-704. Rowe, R.C. 1980. Comparative pathogenicity and host ranges of Fusarium oxysporum isolates causing crown and root rot of greenhouse and field-grown tomatoes in North America and Japan. Phytopathology 70: 1143-1148. Rowe, R.C., Farley, J.D. and Coplin, D.L. 1977. Airborne spore dispersal and recolonization of steamed soil by Fusarium oxysporum in tomato greenhouses. Phytopathology 67: 1513-1517. Sanchez, L.E., Endo, R.M. and Leary, J.V. 1975. A rapid technique for identifying the clones of Fusarium oxysporum f. sp. lycopersici causing crown and root rot of tomato. Phytopathology 67: 726-727. Sato, R. and Araki, T. 1974. On the tomato root-rot disease occurring under vinyl-house conditions in southern Hokkaido. Annu. Rep. Soc. Plant Prot. North Japan 25: 5-13. Scher, F.M. and Baker, R. 1980. Mechanism of biological control in a Fusarium-suppressive soil. Phytopathology 70: 412-417. Scher, F.M. and Baker, R. 1982. Effect of Pseudomonas Dutida and a synthetic iron chelator on induction of soil suppressiveness to Fusarium wilt pathogens. Phytopathology 72: 1567-1573.  123  Schneider, R.W. 1984. Effects of nonpathogenic strains of Fusarium oxysporum on celery root infection by F. oxysporum f. sp. apii and a novel use of the Lineweaver-Burke double reciprocal plot technique. Phytopathology 74: 646-653. Schroth, M.N. and Hancock, J.G. 1982. Disease-suppressive soil and root-colonizing bacteria. Science 216: 1376-1381. Shoun, H. and Tanimoto, T. 1991. Denitrification by the fungus Fusarium oxysporum and involvement of cytochrome P-450 in the respiratory nitrite reduction. J. Biol. Chem. 266: 11078-11082 Shoun, H., Suyama, W. and Kim, D. 1991. Unique nitrate/nitriteinducible cytochrome P-450 in Fusarium oxvsporum and related fungal species. Agric. Biol. Chem. 55: 593-596. Simeoni, W.L., Lindsay, W.L. and Baker, R. 1987. Critical iron level associated with biological control of Fusarium wilt. Phytopathology 77: 1057-1061. Sivamani, E. and Gnanamanickam, S.S. 1988. Biological control of F. oxysporum f. sp. cubense in banana by inoculation with Pseudomonas fluorescens. Plant and Soil 107: 3-9. Sivan, A. and Chet, I. 1989. The possible role of competition between Trichoderma harzianum and F. oxysporum on rhizosphere colonization. Phytopathology 79: 198-203. Sivan, A., Ucko, 0. and Chet, I. 1987. Biological control of Fusarium crown rot of tomato by Trichoderma harzianum under field conditions. Plant Disease 71: 587-592. Sneh, B., Dupler, M., Elad, Y. and Baker, R. 1984. Chlamydospore germination of Fusarium oxysporum f. sp. cucumerinum as affected by fluorescent and lytic bacteria from Fusarium-suppressive soil. Phytopathology 74: 1115-1124. Sonneveld, C. and Strayer, N. 1989. Nutrient solutions for vegetables and flowers grown in water or substrates. Serie Voedingsoplossingen Glastuinbouw 8: 10-11. Sonoda, R.M. 1976. The occurrence of a Fusarium root rot of tomatoes in South Florida. Plant Dis. Reptr. 60: 271-274. Steel, R.G. and Torrie, J.H. 1980. Principles and Procedures of Statistics - A Biometrical Approach. Second Edition. McGraw-Hill, New York. 633 pp. Taylor, G.S. and Parkinson, D. 1961. The growth of saprophytic fungi on root surfaces. Plant and Soil 15: 261-267.  124  Thorpe, H.J. and Jarvis, W.R. 1981. Grafted tomatoes escape Fusarium foot and root rot. Can. J. Plant Sci. 61: 1027-1028. Tomsette, A.B. 1989. The genetics and biochemistry of nitrate assimilation in ascomycete fungi. In. Soddy, L., Marchant, R. and Read, D.J. (Eds.) Nitrogen, Phosphorus and Sulfur Utilization by Fungi. Symp. of Brit. Mycol. Soc. Cambridge University Press, Cambridge. pp. 34-57. Tousson, T.A. and Nelson, P.E. 1976. A Pictoral Guide to the Identification of Fusarium Species. Penn. State Univ. Press, University Park. 43 pp. Van Peer, R. and Schippers, B. 1992. Lipopolysaccharides of plant-growth promoting Pseudomonas sp. strain WCS417r induce resistance in carnation to Fusarium wilt. Neth. J. Pl. Path. 98: 129-139. Weller, D.M. 1988. Biological control of soil-borne plant pathogens in the rhizosphere with bacteria. Ann. Rev. Phytopathol. 26: 379-407. Wymore, L.A. and Baker, R. 1982. Factors affecting crossprotection in control of Fusarium wilt of tomato. Plant Disease 66: 908-910. Yuen, G.Y., Schroth, M.M. and McCain, A.H. 1985. Reduction of Fusarium wilt of carnation with suppressive soils and antagonistic bacteria. Plant Dis. 69: 1071-1075.  125  APPENDIX A  126  Table Al. Analysis of variance of disease severity ratings of tomato seedlings inoculated with nit mutant or wild-type Fusarium strain B1 and then with Fusarium oxysporum f.sp. radicislycopersici (FORL) df^ Sum of Squares ^ ^ Source Expt. 1^Expt. 2 Expt. 1^Expt. 2 Total^11^29^7.96^93.4 (% total SS) C1:H z 0 vs. others C2:nit B1 vs. B1 error  77**^95** 0^1  1 1  9^27^23^4  ** = significant at a=0.01  127  ^  Table A2. Analysis of variance of regression of Fusarium oxysporum f. sp. radicis-lycopersici (FORL) inoculum density and disease severity ratings ^ Sum of Squares df^ Source Total linear hyperbolic residual  119^ 1^ 1^ 117^  ** = significant at a=0.01  128  212.98 (% total SS) 44** 25** 31  Table A3. Analysis of variance of disease severity ratings of tomato seedlings inoculated with Fusarium strain nit B1 and then with Fusarium oxysporum f.sp. radicis-lycopersici (FORL) after increasing time intervals df Source  ^  Total^  Sum of Squares  Expt. 1^Expt. 2 99  nit B1 FORL inoc. timing: Cl:-4 vs. 0+3+10+20 C2: 0 vs. 3+10+20 C3: 3 vs. 10+20 C4: 10 vs. 20 nit B1 x FORL timing: nit B1 x Cl nit B1 x C2 nit B1 x C3 nit B1 x C4 ^ error (I) 40 rating date rating date x nit B1 rating date x FORL timing: rating date x Cl rating date x C2 rating date x C3 rating date x C4 rd x nit B1 x FORL timing: rd x nit B1 x Cl rd x nit B1 x C2 rd x nit B1 x C3 rd x nit B1 x C4 ^ error (II) 40  199 1  Expt. 1  Expt. 2  710.0 355.3 (% total SS) 50** 25**  1 1 1 1  4* 8**  5** 9**  1 1 1 1  5**  12** 10**  90  23  5  1 1  14** 2**  2** <1*  1 1 1 1  4** 1*  1** 2**  2**  1**  9  3  1 1 1 1 90  * = significant a=0.05; ** = significant at a=0.01.  129  ^  Table A4. Analysis of variance of regression of Fusarium strain nit B1 inoculum density and disease severity ratings, in the presence of Fusarium oxysporum f. sp. radicis-lycopersici (FORL) ^ df^Sum of Squares Source Total  99^ 1^ 1^  linear hyperbolic residual  97^  ** = significant at a=0.01  130  301.79 (% total SS) 59** 8** 33  Table A5. Analysis of variance of disease severity ratings of tomato seedlings inoculated with Fusarium strain nit B1 and then with Fusarium oxysporum f.sp. radicis-lycopersici (FORL) at increasing time intervals (0-4 days) Source Total nit B1 concentration: Cl: 0 vs. 10+100 C2: 10 vs. 100 FORL inoculation timing: linear quadratic cubic quartic nit B1 x FORL timing: Cl vs. linear Cl vs. quadratic Cl vs. cubic Cl vs. quartic C2 vs. linear C2 vs. quadratic C2 vs. cubic C2 vs. quartic error (I) rating date rd x nit B1 conc.: rd x Cl rd x C2 rd x FORL timing: rd x linear rd x quadratic rd x cubic rd x quartic rd x nit B1 x FORL timing: error (II)  Sum of Squares  df  479.9 (% total SS)  299 1 1  13** 15**  1 1 1 1  7** 9** 4** 2**  1 1 1 1 1 1 1 1  3** 5** 2** 1* 4** 4**  135  23  1  <1**  1 1  <1** <1**  1 1 1 1 8  <1** <1** n/s 5  135  * = significant at a=0.05 ;^** = significant at a=0.01  131  Table A6. Analysis of variance of disease severity ratings of tomato seedlings inoculated simultaneously with Fusarium strain nit B1 and Fusarium oxysporum f.sp. radicis-lycopersici (FORL) at different ratios Sum of Squares  df Source Total  99  19  nit B1:FORL ratio: linear quadratic cubic quartic error (I)  Expt. 2  Expt. 1  45  rating date rating date x ratio: rd x linear rd x quadratic rd x cubic rd x quartic error (II)  5  75** 4**  14  1  3**  1 1 1 1  1** <1**  45  * = significant at a=0.05; ** = significant at a=0.01.  132  Expt. 2  260.4 43.6 (% total SS) 86** 7** 2*  1 1 1 1 15  Expt. 1  3  Table A7. Analysis of variance of disease severity ratings of plants inoculated simultaneously with Fusarium strain nit B1 and Fusarium oxysporum f.sp. radicis-lycopersici (FORL) at different total inoculum densities Source Total experiment ratio total density: linear quadratic expt. x ratio expt. x total density: expt. x linear expt. x quadratic ratio x total density: ratio x linear ratio x quadratic expt. x ratio x density: expt. x ratio x linear expt. x ratio x quadratic error  df  Sum of Squares  119  89.17 (% total SS) 2 10**  1 1 1 1 1  <1 <1 4*  1 1  <1 1  1 1  <1 2  1 1  1 <1  108  80  * = significant at a=0.05; ** = significant at a=0.01  133  Table A8. Analysis of variance of colony forming units of Fusarium strain nit B1 or Fusarium oxysporum f.sp. radicislycopersici recovered from previously-inoculated tomato seedling roots Experiment Source^ 1  2  ^  ^  Total^  29^831,702 % total SS fungal strain^1^54** nit B1 conc.^1^<1 strain x nit B1 conc.^1^<1 ^ error 26^46  15^2448 (% total SS) fungal strain^1^ 3  Total^  error^  3  ^  df^Sum of Squares  14^97  Total^  15^179,960 (% total SS) fungal strain^1^ 11  error^ ** = significant at a=0.01  134  14^ 89  Table A9. Analysis of variance of regressions of colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f.sp. radicis-lycopersici (FORL), and time (short-term) ^ df^Sum of Squares Source  Nit B1 Total^ linear^ quadratic^ cubic^ quartic^ ^ residual  FORL (days 11-14) Total^ linear^ ^ residual  31^ 21,063,606 (% total SS) 41** 1^ 4 1^ <1 1^ 1 1^ 27^  54  24,575 19^ (% total SS) 2 1^ 18^  ** = significant at a=0.01  135  98  Table A10. Analysis of variance of regressions of colony forming units of Fusarium strain nit B1 and Fusarium oxysporum f.sp. radicis-lycopersici (FORL), and time (long-term) Source  df  Sum of Squares  Surface Colonization: Nit B1 Total  79  2,956,384 (% total SS) 86** 1 <1 <1  linear quadratic cubic quartic  1 1 1 1  residual  75  Surface Colonization: FORL Total  44  linear quadratic cubic quartic  1 1 1 1  residual  40  From Ground Roots: Nit B1 Total  79  linear quadratic cubic quartic  1 1 1 1  residual  75  From Ground Roots: FORL Total  44  linear quadratic cubic quartic  1 1 1 1  residual  13 1,942,130 (% total SS) 28** <1 7* 4 60 532,052 (% total SS) 69** 4** 2* <1 25 22,655 (% total SS) 33** 5 <1 <1  40  * = significant at a=0.05; ** = significant at a=0.01  136  62  Table All. Analysis of variance of mean disease severity ratings of plants treated with broth culture filtrates of Fusarium strain nit B1 and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL) ^ df^Sum of Squares Source Total^  35^112 (% total SS) 1^ 77**  FORL^ broth treatment: Cl: living nit B1 vs. broths ^1^ 8** C2-05: variation among broths ^4 FORL x broth treatment: FORL x Cl^ 1^ 13** FORL x C2 - FORL x C5^4 error^  24^  ** = significant at a=0.01  137  2  Table Al2. Analysis of variance of disease severity ratings of plants treated with sterile exudates from Fusarium strain nit B1infested roots and then with Fusarium oxysporum f. sp. radicislycopersici (FORL) ^ df^Sum of Squares Source 39^56.5 (% total SS) 1^64**  Total^  FORL^ elicitor type: Cl: non-sterile vs. sterile^1^14** C2: H 2 O vs. exudates^1^ 1* C3, C4: variation within exudates 2 FORL x elicitor type: 1^15** FORL x C1^ 1^ 1* FORL x C2^ FORL x C3, FORL x C4^2 ^ 30^ 4 error * = significant at a=0.05; ** = significant at a=0.01  Table A13. Analysis of variance of disease severity ratings of plants treated with a fungal cell wall fraction from Fusarium strain nit B1 or Fusarium oxysporum f. sp. radicis-lycopersici (FORL), and then inoculated with FORL ^ Source df^Sum of Squares 15^3.64 (% total SS) strain (nit B1 or FORL)^1^ 1 form (pellet or supernatant) ^1^48** strain vs. form^ 1^ 1 ^ 12^50 error Total^  ** = significant at a=0.01  138  Table A14. Analysis of variance of disease severity ratings of tomato plants inoculated with Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici in the presence of excess glucose [with nitrogen] Sum of Squares  df Source^Expt. 1 Total^79  Expt. 2  Expt. 1  Expt. 2  561.6 224.6 (% total SS) 28** 17** 37** 34** 31* 33**  199  1 nit B1 FORL 1 nit B1 x FORL 1 glucose: 5** lin 1 2** quad 1 <1* res 2 nit B1 x glucose: 1 4** lin 1 2** quad <1* res 2 FORL x glucose 4 nit B1 x FORL x 4 glucose ^ error 60^180^3^4  * = significant at a=0.05; ** = significant at a=0.01 Table All.  139  Table A15. Analysis of variance of disease severity ratings of tomato seedlings inoculated with Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici in the presence of excess iron ^ df^Sum of Squares Source Experiment 1 ^ Total  59^43 (% total SS) 1 1^8*  FORL^ Fe 3+ source^ Fe level: 1^13** linear^ 1^4 quadratic^ FORL x Fe 3+ source^ 1^1 FORL x Fe level: 1 FORL x linear^ FORL x quadratic^ 1 Fe 3+ source x Fe level: source x linear^ 1^7* source x quadratic^1^2 FORL x Fe 3+ source x Fe level: FORL x source x linear^1^3 FORL x source x quadratic^1 error^  48^60  Experiment 2 Total^  199^589 (% total SS) 1^33** 1^32**  FORL^ nit B1^ Fe level: 1^<1** linear^ 1^<1** quadratic^ 2 residual^ 1^34** FORL x nit B1^ 4^n/s FORL x Fe level^ B1 x Fe level: 1^<1** B1 x linear^ 1^<1* B1 x quadratic^ 1^<1 B1 x cubic^ 1 B1 x quartic^ FORL x nit B1 x Fe level^4^n/s error^  180^1  * = significant at a=0.05; ** = significant at a=0.01 140  Table A16. Analysis of variance of root weights of tomato seedlings inoculated with Fusarium strain nit B1 and Fusarium oxysporum f. sp. radicis-lycopersici (FORL) in the presence of excess iron ^ Sum of Squares df^ Source Total^  79^  1^ FORL^ 1^ 81^ FeEDTA concentration^1^ 1^ FORL x B1^ FORL x FeEDTA conc.^1^ B1 x FeEDTA conc.^1^ FORL x B1 x FeEDTA conc.^1^ ^ 72^ error  0.0347 (% total SS) 9** 4** 56** 9** 5** 4** 3** 10  ** = significant at a=0.01  Table A17. Analysis of variance of colony forming units of Fusarium strain nit B1 or Fusarium oxysporum f. sp. radicislycopersici (FORL) washed from the roots of tomato seedlings inoculated with these strains in the presence of excess iron ^ Sum of Squares Source df^ Total^  39^  strain (nit B1 or FORL)^1^ FeEDTA concentration: 1^ linear^ quadratic^ 1^ cubic^ 1 strain x FeEDTA conc.^3^ error^  32^  ** = significant at a=0.01  141  894967 (% total SS) 36** 6 1 2 55  Table A18. Analysis of variance of disease severity ratings of plants treated with the iron chelator EDDHA and then inoculated with Fusarium oxysporum f. sp. radicis-lycopersici (FORL) Source  df^Sum of Squares  Total  79  FORL chelator: complexed or not chelator conc.: linear quadratic cubic FORL x chelator FORL x level chelator x level FORL x chelator x level: FORL x chelator x linear FORL x chelator x quad. FORL x chelator x cubic error  1 1  251 (% total SS) 91**  1 1 1 1 3 3  1**  1 1 1  1**  64  7  * = significant at a=0.05; ** = significant at a=0.01  142  APPENDIX B  143  MEDIA RECIPES 1. Fusarium Minimal Medium (FMM) (modified from Hadar et al., 1989) g/L Sucrose^  30  NaNO 3^2 KH2 PO 4^ ^ MgSO 4 •7H 2 0 ^ KC1  1  0.5 0.5 20  agar^ Add 0.2 ml of trace element solution to 1 L of medium from above trace element solution:  g/95 ml Citric acid^  5  ZnS0 4 '7H2 0^  5  FeSO4^  4.75  Fe(NH 4 ) 2 (SO 4 ) 2 . 6H 2 0)^  1  CuS0 4 .5H 2 )^  250 mg  MnSO 4 • H 2 0)^  50 mg  H3 BO 3^  50 mg  Na 2MOO 4 '2H 2 0^  50 mg  1.8 ml/L of 25% lactic acid may be added to reduce the risk of bacterial contamination 2. Chlorate - Minimal Medium (FMM-C) Minimal Medium as above, with 15 g/L KC1O 3 and 1.6 g/L Lasparagine.  144  3. Media for Distinguishing Nit Mutant Phenotypes (modified from Correll et al., 1987) Basal Medium g/L Sucrose^  30  KH 2 PO4^ ^ MgSO 4 -7H2 0 ^ KC1 ^ agar  1  0.5 0.5 20  Add 0.2 ml/L trace element solution to 1 L basal medium trace element solution: As in Fusarium Minimal Medium (previous page) Nitrate Medium: basal medium + 2 g/L NaNO 3 Nitrite Medium: basal medium + 0.5 g/L NaNO 2 Hypoxanthine Medium: basal medium + 0.2 g/L hypoxanthine Ammonium Medium: basal medium + 1 g/L ammonium tartrate Uric Acid Medium: basal medium + 0.2 g/L uric acid  145  4. Fraser Valley Greenhouse Grower's Fertilizer Solution (B. C. Ministry of Agric. and Fisheries, 1987) g/L Ca(NO 3 ) 2 . 4H2 0^  0.76  KNO 3^0.55 MgSO 4 •7H 2 0^  0.25  KH 2 PO 4^  0.28  Add 1 ml of trace element solution to 1 L fertilizer solution from above trace elements solution:^g/L MnS0 4 • H2 O  1.54  ZnS0 4 ' 7H 2 0  0.44  Na 2M00 4 • 2H 2 0  0.128  CuSO 4  .  5H 2 0  0.12 2.86  H3 B0 3  "Sequestrene" (iron chelate)  15.0  5. Hoagland Solution (Hoagland and Arnon, 1938) ml in 1 L solution 0.1 M KH 2 PO4^  10  0.1 M KNO 3^  50  0.1 M Ca(NO 3 ) 2^50 0.1 M MgSO4^20 Add 1 ml of trace element solution to 1 L fertilizer solution trace element solution: As in Grower's Fertilizer Solution (above) 146  6. Dutch Fertilizer Solution (Sonneveld and Strayer, 1989) g/L Ca(NO 3 ) 2 . 4H 2 0)^  0.649  MgSO 4 7H 2 0^  0.246  KH 2 PO4^  0.170  NH4 NO3^  0.080  KNO 3^  0.430  K 2 SO 4^  0.085  Add 1 ml of trace element solution to 1 L fertilizer solution from above trace element solution: g/100 ml MnSC)4 . H 2 O  0.169  ZnS0 4 • 7H 2 0  0.115  H3 B0 3  0.124  CuSO 4 • 5H 2 )  0.0187  Na 2MOO 4 .2H 2 0  0.0121  Add 1.5 ml iron solution to 1 L fertilizer solution from above iron solution: Na 2 EDTA  0.373  FeS0 4 • 7H 2 0  0.278  (dissolve EDTA first using gentle heat (50 °C) then add Fe salt) Adjust pH to 5.5 - 6.5 using KOH  147  APPENDIX C  148  PHYSIOLOGICAL PHENOTYPES OF NIT MUTANTS  Methods  The physiological phenotypes of nit mutants of three nonpathogenic fusaria (B1, IPA and FS) were determined according to the method of Correll et al. (1987). Five selective media were prepared which consisted of a basal medium (see recipe in Appendix B) supplemented with one of five nitrogen sources: nitrate (2 g/1 NaNO 3 ), nitrite (0.5 g/1 NaNO 2 ), hypoxanthine (0.2 g/1 hypoxanthine), ammonium (1 g/1 ammonium tartrate) or uric acid (0.2 g/1 uric acid). A small piece of mycelium from each of the nit mutants, as well as wild type B1 (as a control) was  °  placed on each plate and these were incubated at 22 C in the dark. After 4 days, each colony was scored for growth relative to the wild-type strain. The nit mutants were also tested for nitrite excretion. Each nit mutant, and each corresponding wildtype strain was grown alone on urea medium (basal medium + 0.4 g/1 urea) for 4 days. Then the plates were flooded with 10 ml of a 3M solution of NaNO 3 and left for 24 hours. After this 1 ml of a sulfanilamide solution (75 ml distilled H 2 O, 25 ml concentrated HC1, 1 g sulfanilamide) and 1 ml of a colour indicator (100 ml distilled H 2 O, 20 mg N-1-napthyl-ethylenediamine•2HC1) were dropped onto the colonies, or, for a control, onto an empty plate of nitrite medium. The presence of nitrite turns the colony bright fuschia. 149  Results and Discussion  Correll et al. (1987) determined the physiological phenotypes of a range of Fusarium oxysporum nit mutants on the basis of knowledge about nitrogen assimilation pathways in  Aspergillis nidulans and Neurospora crassa. The hypothetical pathway in these fungi is shown in Figure Cl. Mutants nit B1 and nit IPA grew on a medium that had nitrite as the sole nitrogen source (Table C1) and so were thought to have mutations in the locus that codes for nitrate reductase. Nit B1 did not grow on nitrate medium and so was thought to have a mutation in the pathway-specific regulatory locus. The product of this gene induces the transcription of structural genes of nitrate reductase and nitrite reductase with a feedback mechanism: when nitrate is not present, the nitrate reductase molecule is thought to convert this gene product to an inactive state in which it can no longer stimulate transcription (Tomsette, 1989). Since all the nit mutants generated for this study grew on hypoxanthine and uric acid, mutations in the major nitrogen regulatory locus or in the molybdenum-containing cofactor (which assembles nitrate reductase into an active enzyme) were not indicated. The product of the major nitrogen regulatory locus represses synthesis of both nitrate reductase and nitrite reductase when a preferred nitrogen source is present.  150  Table Cl. Identification of nitrate non-utilizing (nit) mutants of nonpathogenic Fusarium strains by their growth on different nitrogen sources Growth on nitrogen sources a nit- nit- ammo- hypox- uric ^Nitrite^Probable Strain^rate rite nium anthine acid^excretionb Mutation`  nit B1^-^-^+^+^+^no^pathway-  specific regulatory locus  nit IPA^—^+^+^+^+^no^nitrate  reductase structural locus  nit FS^—^+^+^+^+^no^nitrate reductase structural locus  wt B1^+^+^+^ (Control) a Growth  slight  ^  none  on basal medium with various nitrogen sources: + = typical wild-type growth; - = thin spreading growth with no aerial mycelium bAs described by Cove (1976) no = no colour change; slight = light pink reaction `Following the interpretation of Correll et al. (1987), based on analysis of mutants from Aspergillis nidulans and Neurospora crassa  151  hypoxanthine ^ xanthine ^  )uric acid  purine dehydrogenase ^ molybdenum-containing co-factor  major nitrogen regulatory locus  nitrate ^ nitrite ^ ammonium nitrate^nitrite reductase^reductase  pathway-specific regulatory locus  Figure Cl. Nitrogen utilization pathway in Neurospora crassa and Aspergillis nidulans (from Correll et al., 1987)  152  In the time since these experiments were done, new evidence has been presented that suggests that the nitrogen assimilation pathway in Fusarium oxysporum may actually be quite different from that found in Aspergillis nidulans and Neurospora crassa. This new research showed that Fusarium oxysporum was capable of reducing nitrate or nitrite to a gaseous form of nitrous oxide (N 2 0). The pathway functioned even when a preferred nitrogen source was present, but was suppressed by high aeration (Shoun et al., 1991). The important discovery was that this process involved a cytochrome P-450, which seemed to function as a dissimilatory nitrite reductase. This protein was specifically induced by nitrate or nitrite (Shoun and Tanimoto, 1990). In some preliminary experiments, nit IPA and nit FS showed greater biological control ability than their wild-type counterparts (data not shown), so the nit mutation may be significant, and its effect should be clarified if those strains are to be used in future biological control experiments. Nit B1 was used in the main experiments for this research, and in this strain the mutation had no apparent effect on biological control.  153  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086176/manifest

Comment

Related Items