fiffiw» . $351.13 a ..r. in. . .anfluefigwi 11. a . E . I a . ‘ .b! ‘l 0-. ‘ 1.12:! I & .3 .1... . J. .31: 53wfli301 . . . 35...... ... . a. .3." . I . L3. In: ..: \ .. 3) «.3 {4.3. 5.. 51...... l)!- u' I . . ‘3 I. u 1‘ I . , . , V - 1.... 1L. . 52.1..» . a )l!..;. 1.5.52}. :1. it 9.2.1: i ..e? I 33!...125 :2 ‘ is... ; 3.6.... 5:12;, . 2:141:13. .f. .a. i :3 hr. . “Unix. ,. hufi ...« yu‘hu .5 z.‘ . I MEMLK . . . .y.f‘¢r:.V 1.0 Ti. .1112132. 0.15.. 7 , ...;-..z..1u. “wimp,“«nfi. . r3» . .$m..u....fi.,>fi.. a? ; :fififi. 3m .35 .w.‘.,,.,.u4.2.ar.§. .1 #4 .. . lllllllllll’HlllHllllilllllhllllll‘thllllllHlllHllll 31293 01016 2760 This is to certify that the thesis entitled Examination of Genetic and Cultural Controls for Fusarium Yellows of Celery presented by Elizabeth Jean Hudgins has been accepted towards fulfillment of the requirements for MS degree in Want Pathology 93W if or professor Date /g//é—/74]Z 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY l Michigan State ,3 University i PLACE ll RETURN BOX to man this checkout tram your "cord. TO AVOID FINES rattan on or baton dd- duo. DATE DUE DATE DUE DATE DUE i—I—J MSU loAnNflnnutivo Adlai/Equal Opportunity initiation Wan-m EXAMINATION OF GENETIC AND CULTURAL CONTROLS FOR FUSARIUM YELLOWS OF CELERY By Elizabeth Jean Hudgins A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Botany and Plant Pathology 1 994 ABSTRACT EXAMINATION OF GENETIC AND CULTURAL CONTROLS FOR FUSARIUM YELLOWS OF CELERY By Elizabeth Jean Hudgins Fusarium yellows of celery, caused by Fusarium oxysporum f.sp. apii race 2, is the most destructive disease of celery in Michigan for which effective controls are not available. Celery somaclones derived from the moderately resistant cultivar Tall Utah 52-70 HK and the highly susceptible cultivar Florida 683 were screened annually for Fusarium yellows resistance and field selections were made. Tall Utah 52-70 HK-derived somaclonal lines showed improvement in resistance compared to Tall Utah 52-70 HK; germplasm release is set for 1995. The Florida 683-derived somaclones also were more disease resistant than Florida 683; selections will continue for 2 or 3 years. Furanocoumarins inhibited growth of Fusarium oxysporum f.sp. apii race 2 with or without ultraviolet irradiation. Since furanocoumarin content was substantially elevated only in the FL 1-2 somaclonal line, it was unlikely that furanocoumarin expression was responsible for the Fusarium yellows resistance observed. Celery, onion, rye, cabbage, rape, or mustard did not reduce Fusarium yellows incidence in greenhouse studies when used as soil amendments or cover crops. Tarping and heating soil amended with onion or mustard significantly lowered disease ratings of Fusarium yellows of celery. ACKNOWLEDGMENTS I would like to thank Dr. Melvyn Lacy, my major professor, for his encouragement, advice, and assistance in preparation of this thesis. He provided me with many rewarding opportunities in research, extension, and professional development. I am appreciative of the financial support from Mel Lacy for the duration of this project which was funded in part by Celery Research Inc. , Hudsonville, MI and USDA/CSRS special grants. I am also grateful to my committee members, Drs. Rebecca Grumet, Ray Hammerschmidt, and Pat Hart, for research suggestions and encouragement. Special thanks goes to Brian Cortright for his tireless help with my research. Thanks are also extended to many persons in the Department of Botany and Plant Pathology at Michigan State University. In particular, I acknowledge Dave Huber for his friendship and supportive interest in my work, Jeff White for his kindred spirit and helpful inquiries which enabled me sort out my thoughts, Dave Jacobson for his friendship and care which helped me grow professionally and personally, and to Lissa Leege for her enthusiasm and companionship that helped my last two years come alive. iii CEI TABLE OF CONTENTS Page LIST OF TABLES vi LIST OF FIGURES viii LIST OF ABBREVIATIONS x CHAPTER 1. LITERATURE REVIEW Celery 1 Fusarium oxysporum f. sp. apii history and geographical distribution 3 Profile of celery pests and disorders 5 Fusarium yellows disease symptoms 6 Fusarium oxysporum f. sp. apii identification 6 Host and nonhosts - carriers of FOA2 10 Vascular wilt infection and pathogenesis of Fusarium oxysporum 13 Soil ecology of Fusarium oxysporum 15 Control 16 LITERATURE CITED 22 CHAPTER 2. SOMACLONAL VARIATION TO IMPROVE RESISTANCE OF CELERY TO FUSARIUM YELLOWS INTRODUCTION 31 MATERIALS AND METHODS 39 Somaclone planting 39 Harvesting and evaluation 40 Preparation of plants for vemalization 42 Vernalization 43 Greenhouse conditions post-vernalization and seed collection 43 RESULTS 44 Tall Utah 52-70 HK-derived somaclone screening 44 Florida 683-derived somaclone screening 51 DISCUSSION 55 LITERATURE CITED 62 iv CHAPTER 3. EVALUATION OF SOIL AMENDMENTS AND COVER CROPS FOR THEIR EFFECT ON DISEASE EXPRESSION INTRODUCTION 66 MATERIALS AND METHODS 78 Inoculum production 78 Selection of best soil infestation method 78 Soil amendments 80 Soil heating and mulching 81 Cover crops 82 Statistical analysis 83 FOA2 isolations from celery crowns 83 RESULTS 85 Determination of soil infestation method 85 Effect of organic amendments on Fusarium yellows of celery 92 Effect of cover crops on Fusarium yellows of celery 95 FOA2 isolations from celery crowns 95 DISCUSSION 100 LITERATURE CITED 108 CHAPTER 4. EFFECT OF FURANOCOUMARIN ON FUSARIUM OXYSPORUM F.SP. APII RACE 2 INTRODUCTION 1 15 MATERIALS AND METHODS 120 Dry weight assays 120 Radial growth assays 121 Statistical analysis 121 Furanocoumarin extraction from celery tissue 122 RESULTS 124 Effect of S-MOP on dry weight of FOA2 124 Effect of 5-MOP on radial growth of FOA2 124 Effect of 8-MOP on dry weight of FOA2 124 Effect of 8-MOP on radial growth of FOA2 131 Furanocoumarin content in somaclones 138 DISCUSSION 141 LITERATURE CITED 144 LIST OF TABLES Table 1 4 Reactions of R4 and R5 Tall Utah 52—70 HK-derived celery lines Reactions of bulk Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f.sp. apii race 2 in a replicated field trial near Muskegon, MI, 1992. Reactions of R, Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in an unreplicated field trial near Muskegon, MI, 1992. Reactions of R5 bulk Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f.sp. apii race 2 in a replicated field trial near Hudsonville, MI, 1993. to Fusarium yellows in soil naturally infested with Fusarium oxysporum f.sp. apii race 2 in an unreplicated field trial near Decatur, MI, 1993. Reactions of bulk Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in a replicated field trial near Hudsonville, MI, 1994. Reactions of Florida 683-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f.sp. apii race 2 in an unreplicated field trial near Muskegon, MI, 1992. Reactions of R2 and R3 Florida 683-derived celery lines and controls to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in an unreplicated field trial near Decatur, MI, 1993. vi Page 45 47 49 50 52 53 Table Page 8 Percentage of Fusarium oxysporum f. sp. apii race 2 isolates compared with total Fusarium oxysporum isolated from Florida 683 celery crowns based on colony size on the selective D-medium. 99 9 Estirnations of furanocoumarin content in celery lines derived from somaclonal variation compared with parent cultivars Florida 683 and Tall Utah 52~70 HK, using thin-layer chromatography. 140 vii LIST OF FIGURES Figure 1 Lineage of Florida 683-derived celery somaclones selected for increased resistance to Fusarium yellows of celery. Effect of soil type on mean Fusarium yellows disease ratings and fresh weight of Florida 683 celery seedlings 6 weeks after transplanting into soil artificially infested with low levels of FOA2. Effect of planting depth, inoculum level, and isolate type on the mean Fusarium yellows disease ratings and fresh weight of Florida 683 celery seedlings, 6 weeks after transplanting, using naturally infested soil from Hudsonville, MI. Effects of steaming muck soil and using wet or dry FOA2- colonized wheat straw inoculum on mean Fusarium yellows disease ratings and fresh weights of Florida 683 celery seedlings, 6 weeks after transplanting. Effect of soil amendments on mean Fusarium yellows disease ratings of Florida 683 celery seedlings 8 weeks after transplanting into FOA2-infested soil in the greenhouse. Effect of cover crops on mean Fusarium yellows disease ratings of Florida 683 celery seedlings 8 weeks after transplanting into FOA2-infested soil in the greenhouse. Effect of S-methoxypsoralen (5-MOP), with or without a one hour 366 nm ultraviolet light exposure, on dry weight of Fusarium oxysporum f.sp. apii race 2 after 7 days. Effect of 5—methoxypsoralen (5—MOP), with or without a one hour 366 nm ultraviolet light exposure, on radial growth of Fusarium oxysporum f.sp. apii race 2 after 5 days. viii Page 59 86 88 93 96 126 128 Figure Page 9 Effect of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on dry weight of Fusarium oxysporum f.sp. apii race 2 after 7 days. 130 10 Effect of high doses of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on dry weight of Fusarium oxysporum f.sp. apii race 2 after 7 days. 133 11 Effect of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on radial growth of Fusarium oxysporum f.sp. apii race 2 after 5 days. 135 12 Effect of high doses of 8-methoxysporalen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on radial growth of Fusarium oxysporum f. sp. apii race 2 after 5 days. 137 ix cm 8-MOP S-MOP FOA2 v/v w/w LIST OF ABBREVIATIONS centimeters Celsius 8-methoxypsoralen feet 5-methoxypsoralen Fusarium oxysporum f. sp. apii race 2 grams kilograms liters meters Michigan microliters micrograms milligrams milliliters millimeters millirnolar molar parts per million potato dextrose agar pounds rotations per minute ultraviolet volume for volume weight for weight CHAPTER 1 - LITERATURE REVIEW Celery The earliest documented uses of celery (Apium graveolens L.) were for the making of wreaths by the ancient Greeks (Snowdon 1992) and for treating various ailments including hypertension (Ezzell 1992). The first major commercial use of celery was for opiate production in the early 1800’s (G. Hill, personal communication). In the United States, the fust commercially produced celery, presumably for opiate extraction, came from Kalamazoo, Michigan in 1856 (Berger 1986, Beattie 1907); the celery seed had been brought to the US. by Dutch settlers. At this time, celery was a ”richman’s" commodity. Since celery stalks do not store well for long (Ryder 1979), celery was one of the luxury items that only the rich could afford. Celery became a status symbol of wealth and was displayed in crystal ”celery vases" in the dining rooms of the rich (Edmondson 1992). The invention of the refrigerated railcar in the 1870’s, which used cntshed ice, enabled vegetable shipment without rapid decay (Edmondson 1992); celery was no longer a symbol of affluence. Refrigeration also provided celery shipment from the warmer growing areas to the northern markets, so that by the late 1800’s, celery was mainly being produced in California (G. Hill, personal communication) and Florida (Berger 1986), which is still true today (USDA 1993). Celery is thought to have been domesticated from a wild celery like smallage (G. Hill, personal communication). Although first used for medicinal purposes, celery is 2 now used foremost as a vegetable. Early breeding improvements such as enlargement of the petioles and reduction of the bitter alkaloids and strong odor are thought to have been critical for palatability (Ryder 1979). Commercial celery has been divided into three main types that refer to their color upon blanching: Easy Blanching Green (which blanches white), Yellow, and Green (G. Hill, personal communication). Presently, the Green types are the most commonly grown in the USA because of better Fusarium resistance (G. Hill, personal communication) and consumer favoritism. Prior to the second World War, celery blanching (bleaching of stalks by protecting them from light) was common practice (Beattie 1907). The onset of the war reduced labor availability and celery blanching was discontinued (G. Hill, personal communication). Of the remaining Green types, there are 3 major groups of celery: Pascal, Utah, and Crystal Jumbo. Many of the varieties traditionally grown in Michigan were thought to be from the Crystal Jumbo group (G. Hill, personal communication). Celery has been in the news recently for the studies that have shown that celery may have cancer-fighting potential (Zheng et al. 1993, Sobti et al. 1991) as well as the ability to lower blood pressure (Ezzell 1992). The active compound 3-n-butyl phthalide was found to both reduce levels of catecholamines - thereby relaxing blood vessels - and to induce the production of an enzyme that detoxifies a cancer-causing agent (Zheng et al. 1993). Celery is high in water (95%) and fiber (0.7 % w/w) and low in calories (9 kCal/large stalk) which has made it a diet food (Lorenz and Maynard 1988). In 100g, there are 0.7g protein, 0.1g fat, 3.6g carbohydrate, 6.3 mg vitamin C, 127 IU vitamin A, 0.03mg Thiamine, 0.03mg Riboflavin, 0.3mg Niacin, 0.03mg vitamin B6, 88 mg 3 sodium, 284 mg potassium, 26mg phosphorus, 0.5mg iron, and 36 mg calcium (Lorenz and Maynard 1988). Celery may be one of the most labor intensive of the major vegetable crops to produce, averaging 273-302 man-hours per acre (Berger 1986) if harvested manually. Most celery growers in Michigan now use mechanical harvesters and planters which reduces the estimate. Celery, however, yields $7-12 per crate, 700-900 crates per acre, which makes it a very profitable crop (MDA 1993). Michigan is the third largest state for celery production (2,900 acres, $12.5 million in 1991) following California (20,900 acres, $ 144 million in 1991) and Florida (7,400 acres, $ 37.6 million in 1991) (USDA 1993). Other locations for production in the United States are New York, New Jersey, Ohio, Utah, Pennsylvania, Washington, Wisconsin, and Massachusetts (USDA 1993). Rtsarium oxysporum f.sp. apii History and Geographical Distribution The soil-borne fungus Fusarium oxysporum f.sp. apii (R. Nels. & Sherb.) Snyd. and Hans. (referred to as FOA) incites Fusarium yellows disease in celery (Apium graveolens L. var dulce (Mill.) Pers.) (Hart and Endo 1978). Although Ryker (1935) believed that Fusarium oxysporum f.sp. apii (race 1) first appeared in 1909 in Kalamazoo, Michigan, Coons (1915) was the first to report Fusarium yellows of celery in Michigan which he called "stunting disease of celery". This disease has also been referred to as "sickness", ”root rot", and "crown rot" (Nelson et al. 1937). Fusarium yellows of celery was an economically important disease to celery growers (except in Florida) in the second and third decades of this century (Nelson et al. 1937), causing losses of $200,000/year (1931 value) in Michigan alone (Nelson et al. 4 1937). Green cultivars provided better disease resistance, and replaced the self-blanching cultivars grown before 1939 (Newhall 1953). In 1952, the green cultivar Tall Utah 52- 70 was found to be highly resistant to FOA (Elmer 1985); it replaced most cultivars grown in the United States. Tall Utah 52-70 was so resistant that FOA was essentially eliminated from the soils in the United States (Elmer 1985). The high level of disease control maintained with the Tall Utah 52-70 and other green cultivars explains the dearth of literature and research on Fusarium yellows of celery after 1952 until the disease resurfaced in 1975 in California (Hart and Endo 1975). The causal agent of the most recent outbreak of Fusarium yellows was shown to be a new race, designated Fusarium oxysporum f.sp. apii race 2 (FOA2) (Schneider and Norelli 1981). The new race was pathogenic not only to the self-blanching, yellow cultivars but also to the previously resistant green cultivars, including Tall Utah 52-70. The pathogen causing the earlier disease of yellow cultivars was designated Fusarium oxysporum f. sp. apii race 1 (FOAl) (Schneider and Norelli 1981). Puhalla (1984b) reported a third race, FOA race 3, that only attacked green cultivars of celery, although this finding was not confirmed. N 0 reports of FOA race 3 outside of California have been published. Puhalla ( 1984b) speculated that races 1 and 2 are unrelated but that races 2 and 3 are closely related, possibly race 3 arose from race 2. The spontaneous arrival of Fusarium races are unpredictable. A single mutation may change an avirulent strain to a virulent one (Burnett 1984); this has been proposed to be the case for the sudden appearance of FOA2 (Toth 1989). FOA race 2 may have originated in California (Toth 1989) and its rapid spread across much of the celery growing areas in the United States and into Canada may have 5 been through infested soil, contaminated farm machinery, and celery transplants that carried FOA2 propagules (Elmer 1985). The presence of FOA2 has been confirmed in California (Hart and Endo 1975,1978), Michigan (Elmer and Lacy 1982), New York (Awuah et al. 1986), Texas (Martyn et al. 1987), Ontario (Cerkauskas and McDonald 1989), and British Columbia (Gaye et al. 1991). Its presence is suspected in Florida (R.D. Martyn unpublished). Profile of Celery Pests and Disorders Fusarium Yellows (caused by Fusarium oxysporum f. sp. apii (Nelson & Sherb.) Snyder & Hans.) is a major disease of celery grown in Michigan because it cannot be controlled economically with fungicidal sprays or soil fumigants (M.L. Lacy personal communication). Lacy and Grafius (1980) have summarized celery pests and diseases in Michigan. Other fungal diseases are late blight (Septon'a apiicola Speg.), early blight (Cercospora apii Fresen.), damping off and crater rot (Rhizoctonia solani Kiihn), and pink rot (Sclerotinia sclerotiorum (Lib.) de Bary). Non-fungal diseases are bacterial blight (Pseudomonas apii), cucumber mosaic (Virus), western celery mosaic (Virus), and aster yellows (Mycoplasma Like Organism). The most common abiotic disorders seen are black heart caused by calcium deficiency at the growing tips and cracked stem caused by boron deficiency. Insect problems most coMonly are aphids, variegated cutworms, loopers, and carrot weevils. Fusarium Yellows Disease Symptoms The symptoms of Fusarium yellows of celery are stunting, foliar and petiole chlorosis, orange to brown vascular discoloration, crown rot, and wilting (Awuah et al. 1986, Hart and Endo 1978). Severely infected plants often have an orange-brown or black dry rot which may extend into the petioles. Such plants usually die before harvest and rotted areas in the crown may be colonized by a secondary saprophyte causing additional rot (Nelson et al. 1937, Otto et al. 1976). Wilting and crown rot are generally symptomatic of the last stages of disease and of impending plant death (Ehner 1985). Fusarium oxysporum f. sp. apii Identification Fusarium is a form genus within the Hyphomycetes (subdivision Deuteromycotina) (W indels 1992). Fusaria are divided into sections that group together species with similar characteristics (Nelson et al. 1983). The primary method for the identification of Fusarium species is morphology, particularly the morphology of the macroconidia (Nelson et al. 1983). The high level of genetic variability and phenotypic plasticity in vitro (Armstrong and Armstrong 1968, Nelson et al. 1983, Puhalla 1981) requires standard cultural conditions to identify species using Nelson, Toussoun, and Marasas’ (1983) synoptic keys. Appropriate cultural conditions include growing single-spored cultures on potato dextrose agar and carnation leaf agar at standard light and temperature regimes, and examining at 10-14 days (Windels 1992, Nelson et al. 1983). Colony diameters (W indels 1992), pigmentation (Nelson et al. 1983), colony morphology (Awuah and Lorbeer 1989), as well as microscopic features (conidia, sporodochia, and conidiophores) (W indels 1992, Nelson et al. 1983) are useful for distinguishing species. 7 Although these characteristics may separate species, they are generally unhelpful for the accurate identification of formae speciales. Selective media, physiological, and modern molecular techniques are currently being used for faster, more thorough in vitro identification of Fusaria, including FOA2. A forma specialis is a pathogenic classification not based on morphology; the traditional method for the identification of formae speciales of Fusarium is greenhouse pathogenicity tests (Armstrong and Armstrong 1968, Snyder and Hansen 1940). Virulent isolates of Fusarium oxysporum f. sp. apii have been identified with greenhouse pathogenicity tests by transplanting celery into artificially infested soil (Correll et al 1986a, Elmer 1985, Hart and Endo 1978, Hart and Endo 1981, Puhalla 1984b, Toth and Lacy 1991). These tests may take as long as 13 weeks (Toth and Lacy 1991), produce variable disease ratings (Hart and Endo 1978, Orton 1981, Toth and Lacy 1991) and provide inconclusive results (Correll et al. 1986a, Puhalla 1984b). Environmental conditions may affect the expression of Fusarium yellows of celery, since they have an effect on the pathogenicity of other Fusarium wilts (Cook 1981 , Beckrnan 1987, Dhingra and Sinclair 1985). The poor understanding of the effects of environment on Fusarium yellows disease of celery (Toth 1989) may hamper the ability to perform a reliable virulence test in the greenhouse. Currently, two other main techniques - vegetative compatibility and selective media - are useful for the accurate identification of FOA2. Fungal strains are said to be vegetatively compatible if they are able to anastomose and form a stable heterokaryon (Puhalla 1985). Isolates of Fusarium oxysporum can be tested for their vegetative compatibility with tester strains by producing mutants that, when paired with the tester mutants, will produce wild-type growth after 8 successful heterokaryosis (Correll et al. 1987). Color and auxotrophic mutants have been used with some success (Puhalla 1984a, 1984b, 1985) but nitrate non-utilizing ("nit") mutants have been used most widely (Correll et al 1986a, 1986b, 1987, Ireland and Lacy 1986, Puhalla 1985, Toth and Lacy 1991). Since FOA2 has been determined to belong to a single vegetative compatibility group (V CG) to which no other forma specialis or saprophyte belongs (Toth and Lacy 1991), nit-pairing for the identification of FOA2 is currently one of the most reliable tests that exists for the identification of FOA2. Even so, some of the nit mutants do not form heterokaryons reliably, so that some FOA2 isolates escape correct detection (Correll et al. 1987, Toth and Lacy 1991). Along with vegetative compatibility, another identification technique would help to detect those escapes. Puhalla (1984a) found that L-sorbose and 0.05% (v/v) Tergitol N-lO restricted the growth of F 0A2, which led to the development of a selective medium (DM) for identification of FOA2 (Puhalla 1984b). FOA2 is transferred by stab inoculations (Correll et al 1986a) or individual conidia (Puhalla 1984b) to DM and the colony diameter is measured after 72 hours. Relative colony diameters of the unknown isolates compared with a control isolate, known to be FOA2, are used to identify the unknown as FOA2 or as another forma specialis. Correll et al (1986a) found that all isolates with a relative colony diameter 2 1.5 were not FOA2 and those 5 1.0 were definitely FOA2. Jacobson (unpublished) confirmed these distributions but with the identification of the non-FOA2 isolates at relative diameters of 2 1.3. Puhalla (1984b) had found similar results except for a few FOA2 isolates from fields outside of California that had large colony diameters. Awuah and Lorbeer (1986) also developed a selective medium for FOA2, using L-sorbose as the carbon source. Their sorbose-based medium 9 (SBM) was designed for enumerating FOA2 propagules from organic soils based on the characteristic slow-growing, compact colonies of FOA2. Other attempts have been made to develop techniques to identify FOA2. Toth and Lacy (1990) developed a selective medium (potato carrot agar plus) for the identification of FOA2 nit mutants in artificially infested muck soil. An orange mutant of FOA2 was discovered (Elmer and Lacy 1984a, 1984b, 1987b) that could be easily identified from artificially infested soil, but it had reduced virulence. Awuah and Lorbeer (1989) determined that F 0A2 had two cultural morphologies on chloramphenicol-amended Difco potato dextrose agar (CPDA). Aerial mycelium was produced with dark incubation and pinnotal growth in fluorescent light. Awuah and Lorbeer (1989) did not show that the growth habits differed from any other races of FOA or other formae speciales of Fusarium oxysporum. Electrophoresis has not provided conclusive identification of FOA nor its races. Starch gel electrophoresis was used to separate allozymes of acid phosphatase (Puhalla 1985). Most of the FOA isolates tested had a slow-migrating band of acid phosphatase but a fast-migrating band was associated with all FOA race 1, one FOA race 2 isolate from Taiwan, and two putative FOA race 3 isolates (Puhalla 1985). Separation of total soluble proteins was performed using sodium dodecyl sulfate - polyacrylamide gel electrophoresis (SDS-PAGE) (Toth and Lacy 1991). Although the banding pattern was the same for all FOA2 isolates tested except one, the protein banding pattern could not be distinguished from other Fusarium oxyspomm formae speciales (Toth and Lacy 1991). Currently, vegetative compatibility and colony diameters on DM seem to provide the fastest and most correct identification of FOA2. Other molecular techniques may be 10 developed for a more accurate and rapid diagnosis, especially with the techniques that identify the presence of the pathogen in situ. Restriction fragment length polymorphism (RFLP) analysis has been used with identification of races and formae speciales of Fusarium oxysporum (Kistler et al. 1987, Manicom et al. 1990) but many of the RFLP banding patterns are the same among formae speciales (Correll 1992). Randomly amplified polymorphic DNA (RAPD) analysis, a newer technique, is being examined for its ability to rapidly identify races (Grajal-Martin et al. 1993, Manulis et al. 1994, Assigbetse et a1. 1994) or formae speciales (Manulis et al. 1994) of F. oxysporum. Strain identification of F. oxysporum is also done with DNA probes (Manulis et al. 1994, Kistler et al. 1991) which could be attempted with FOA2. Finally, monoclonal and polyclonal antibodies can be used to identify or distinguish among several Fusaria species (Iannelli et al. 1983), formae speciales (Iannelli et al. 1983), or races (Wong et al. 1988). Most commonly, fungal detection with antibodies is done with ELISA (enzyme- linked immunosorbent assays) (Correll 1992). Antibody production, especially monoclonal, requires an large initial investment of time and labor. However, the screening of isolates with ELISA is straightforward and very precise. FOA2 could potentially be identified using this procedure. Hosts and Nonhosts - Carriers of FOA2 The wilt Fusaria (Fusarium oxysporum) were named for the characteristic damage to the plant (forma suscept) each forma specialis infected, at a time when a narrow host range was imagined (Price 1984). These formae speciales, however, have been isolated from roots of other plants called "symptomless carriers". Such plants do not show any 11 external symptoms of disease but they either are actively invaded by the pathogen or the fungus remains in the rhizosphere (Price 1984, Armstrong and Armstrong 1948); carriers that are internally colonized are referred to as nonsusceptible hosts, nonsuscepts, or secondary hosts. These nonsuscepts that can act as additional hosts for the pathogen may enable survival of the pathogen even in absence of the host by providing nutrients and escape from predators (Curl 1988). Plants that are not colonized by the pathogen (nonhosts) may hamper - at least not promote - pathogen survival. Some nonhosts are able to lower Fusarium oxysporum population levels lower than the populations in soils left fallow (Banihashemi 1968). Studies have been done to establish the ability of a nonsusceptible plant to harbor a pathogen in its tissues (Armstrong and Armstrong 1948, Waite and Dunlap 1953, Hendrix and Nielsen 1958, Katan 1971). Katan (1971) has suggested that perhaps the reason most secondary hosts remain symptomless is because they prevent the pathogen from producing a toxin. Armstrong and Armstrong (1966) hypothesize that pathogens on common hosts must all have common genes for pathogenicity. Katan (1971) speculates that the balanced relationship between the secondary host and pathogen could be natural and that the pathogenic relationship between the primary host and pathogen could be a deviation from this normal or natural state. Primary hosts and fungal pathogens are usually identified using Koch’s postulates (Agrios 1988). Secondary hosts are more difficult to identify because no disease results from infection. Usually the isolation of a pathogen from carrier tissue and reisolation after inoculation is sufficient to establish a plant as a secondary, symptomless host; 12 histological evidence of tissue invasion may identify the secondary host range more accurately (Price 1984). A few studies have been done on the host range of Fusarium oxysporum f. sp. apii. In 1962, Armstrong and Armstrong reported the first host other than celery for Fusarium oxysporum f.sp. apii; a wilted Mexican sunflower was infected with FOA. This FOA isolate was identified on the Golden Self Blanching celery cultivar (Armstrong and Armstrong 1966) so that its race is unidentifiable; however, since FOA2 was not identified until 1975 (Hart and Endo 1975, Schneider and Norelli 1981), it is likely that this fungus was race 1. Garden pea also may be a susceptible host of Fusarium oxysporum f. sp. apii (Armstrong and Armstrong 1967). Armstrong and Armstrong (1966) tested the host range of the FOA isolate that caused sunflower wilt. No other wilt reactions occurred on the wide host range they tried including other plants from the Compositae family. Elmer and Lacy (1987b) also examined the host range of FOA race 2. Using an orange mutant of FOA2, they were able to quickly identify colonized plant tissue. The host range of Fusarium oxysporum f. sp. apii race 2 was tested on bamyard- grass (Echinochloa crusgalli (L.) Beauv.), smartweed (Polygonum pennsylvanicum L.), purslane (Portulaca oleracea L.), lamb’s-quarters (Chenopodium album L.), onion (Allium cepae L.), carrot (Daucus carota L.), celery (Apium graveolens var dulce L.), lettuce (Lactuca sativa L.), peppermint (Mentha piperita L.), sudax (Sorghum sp.), sweet corn (Zea mays L.), cabbage (Brassica oleracea L.), parsley (Petroselinum crispum (Mill) Nym.), and crabgrass (Digitaria sanguinalis (L.) Scop.). They found that the fungus was able to colonize stems of onion, cabbage and smartweed and it was also recovered from the roots of all the plants tested except parsley and crabgrass. Onions, 13 resistant celery, and lettuce roots were colonized by the orange mutant at low levels. No symptoms were visible on any of the tested plants except celery. It would be interesting to see if FOA race 2 causes the reported sunflower and garden pea wilt. Vascular Wilt Infection and Pathogenesis of Fusarium oxysporum Fusarium oxysporum is the most important Fusarium species causing diseases of economic crops (Booth 1984). Vascular Fusarial wilts cause such large losses because of their intrusive and destructive pathogenesis for which there are few reliable control measures. Some Fusaria enter through wounds on host roots (Beckman 1987), however root wounding does not increase the severity of Fusarium yellows of celery (Hart and Endo 1981). Evidence exists that FOA2 infects celery by direct penetration at the root tips (Hart and Endo 1981), mainly in the zone of elongation just behind the root cap (Jordan et al. 1988). Chlamydospore germination was found to be greatest at 1.3 mm behind root apices (Elmer and Lacy 1987b). Infection of the root tips is sufficient to enable vascular colonization (Becker et al. 1990), which then leads to disease expression. Fusarium oxysporum f. sp. apii race 2 penetrates both intercellularly and intracellularly by mechanical and enzymatic means (Jordan et al. 1988). MacHardy and Beckman (1981) reviewed the disease progression induced by vascular wilt Fusaria (formae speciales of Fusarium oxysporum), which is summarized here. The fungus invades the tracheal vessels of the protostele after intercellular and intracellular growth in the parenchyma tissue and growth through the primary meristem. From the protostele, hyphae grow into protoxylem vessels and spread from vessel to vessel through the pits in the vessel walls. Fusarium colonizes the vascular tissue where growth and l4 sporulation is mainly restricted to the xylem vessels. Here they assimilate mainly cellulosic and pectinaceous substances, although ’ there are dilute amounts of carbohydrates, amino acids, and minerals available in the vascular fluid. Once extensive invasion of primary and secondary xylem elements has occurred, an "explosive" stage of colonization follows which produces abundant numbers of spores (usually microconidia). Rapid colonization of the xylem elements above the hypocotyl occurs soon after this explosive stage. In resistant host cultivars, Fusarium oxysporum penetrates roots but its spread within the xylem is limited to a few vessels or to a localized area. Symptom expression of Fusarial wilts have been attributed to toxin production and/or mechanical plugging of the vascular system. Mycotoxins secreted by Fusarium oxysporum (Lycomarasmin and Fusaric acid) may induce wilting, yellowing, vein- clearing, and necrotic symptoms (Scheffer and Walker 1953, Dimond and Waggoner 1953), although MacHardy and Beckman (1981) claim that there is only weak evidence to suggest a role of toxins in vascular wilt pathogenesis. Vascular blockage from gels, gums, tyloses, and vessel collapse have been found to be associated with wilt, yellowing, stunting, and necrosis as well (MacHardy and Beckman 1981). These occlusions may increase resistance to water flow which in turn produces typical symptoms. Mycelial plugs within vessels have also been hypothesized as obstructors of water flow (Salttink and Dimond 1964) but there are few findings to support this view (MacHardy and Beckman 1981). Water stress does not become critical until the late stages of pathogenesis (MacHardy and Beckman 1981), which explains why wilting is rarely seen in celery infected with FOA2 except on very hot days or during periods of drought, and only when the disease is well-progressed (Elmer 1985). 15 Soil Ecology of Fusarium axysporum Fusarium species have been found around the world and Fusarium wilts occur on crops in most countries (Nelson et al. 1981). However, Fusarium oxysporum, the causal agent of Fusarial wilts, has not been isolated from certain soils. Toussoun (1975) indicated that F. oxysporum formae speciales were generally absent from forest soils but common in cultivated soils. Since F. oxysporum has been found in most soil types, it has been suggested that plant cover and its root system are responsible for the presence or absence of this fungus (T oussoun 1975). Once established in a soil, Fusarium oxysporum formae speciales can survive in a dormant state as chlamydospores in soils or in decaying host tissues for long periods of time (Burgess 1981). Chlamydospores, or resting spores, are thick-walled spores formed between consecutive septae along hyphae or in macroconidia. They have a more simplified internal organization than conidia (Marchant 1984). Chlarnydospores are most important for survival in soil because they enable the fungus to survive conditions detrimental to other fungal structures such as conidia or mycelia (Park 1954). These resting spores are often more durable than perithecia which can have a relatively short period of survival, especially if dried (Booth 1984). Within the last few decades, chlamydospores have been acknowledged as the "individual members of an infection reservoir" (Price 1984) and the primary source of inoculum for the following season. Researchers have realized the potential for disease control if chlamydospore population is lowered (Elmer and Lacy 1987a, 1987b, Huber and Watson 1970, Papavizas 1975, Zakaria 1978); this information is reviewed in Chapter 3. Secondary inoculum reservoirs are the conidia and mycelium left on seed and other plant debris (Park 1959). FOA is 16 also a facultative saprophyte (Opgenorth and Endo 1981) enabling it to colonize dead organic substrates in between host infections. Control Fusarium yellows of celery was devastating celery crops during the initial years of its re-introduction, forcing growers to abandon celery production in certain fields. Early attempts at controlling the disease included chemical controls. Vorlex“ fumigation, Vapam" application with irrigation, and Aliette" sprays were all found to be ineffective for Fusarium yellows control in Michigan (M.L. Lacy, personal communication). Benlate", Topsin", Bravo“, and Truban" fungicidal root dips prior to transplanting were also ineffective controls (Otto et al. 1976). Although fungicides were unsuccessful control agents, methyl bromide fumigation eliminated FOA2 (Awuah and Lorbeer 1991). Methyl bromide fumigation may be useful for greenhouse control of Fusarium yellows of celery (Awuah et al. 1986) but it is deemed economically infeasible for use in the field (Otto et al. 1976) except for very localized areas (Awuah and Lorbeer 1991). Celery was also found to take up so much methyl bromide that the plants were inedible (L.P. Hart personal communication). In addition, methyl bromide labels are being cancelled by EPA. Soil fumigants of 3:1 methyl bromide to chloropicrin, 3:2 chloropicrin to D- D", and DD" alone reduced disease incidence and severity of Fusarium yellows of celery in California but they were not considered economic controls (Otto et al 1976). Poor chemical control pushed researchers to examine biological (Elmer 1985, Elmer and Lacy 1987b, Toth 1989) and cultural (Opgenorth and Endo 1983, Schneider 1985) control measures. 17 One of the most promising potential controls is plant resistance. The discovery of a disease resistant cultivar, Tall Utah 52-70, eliminated Fusarium yellows from celery fields within a few years and the disease was unknown in the United States until its reappearance (Hart and Endo 1978). Celery cultivars have been screened for their level of resistance to FOA2 when planted in naturally and artificially infested soils (Opgenorth and Endo 1979, 1985, Elmer 1985). Although most cultivars are susceptible to Fusarium yellows, some, such as Tall Utah 52-70HK and Deacon, were found to tolerate higher levels of soil infestation while exhibiting few disease symptoms (Elmer 1985, Gaye et al. 1991). Plant resistance has been associated with many cellular responses that Ralton et al. (1987) have summarized as follows: silicon (or silicon-containing material) deposition on/in the plant cell walls adjacent to haustoria-colonized cells, hypersensitive reaction (cell death) of infected tissue and surrounding cells, release of phytoalexins and phenolic compounds, callose appositions, papillae formation, lignification, suberization, and increased production of hydroxyproline—rich cell wall glycoproteins. It is often these early plant responses that facilitate or impede fungal growth and determine the level of resistance of the host plant (Ralton et al. 1987). Jordan et al. (1988, 1989) have been looking at the ultrastructural responses of resistant and susceptible celery cultivars to Fusarium oxysporum f. sp. apii race 2 infection. These studies showed an increase in callose deposition in vascular tissue as fungal colonization progressed (Jordan et al. 1988) and more callose was formed in resistant than in susceptible cultivars. It appears that callose helps to prevent colonization and subsequent disease formation of Fusarium yellows of celery. Phenolic substances also have been found associated with 18 incompatible hosts of FOA2 (Jordan et al. 1989). These phenol-containing bodies, which are seen in micrographs as electron opaque structures, increase in number and size as infection progresses. These phenolic compounds may be furanocoumarins, since it was found that resistant cultivars expressed higher levels of these putative phytoalexins than did susceptible varieties (Cerkauskas and Chiba 1991). Attempts to select for resistance using somaclonal variation have produced encouraging results (Toth 1989) and new cultivars being screened by this method will be set for germplasm release in 1995. A review of this technique used to develop celery lines resistant to FOA2 is given in Chapter 2. There have been many other suggestions made about potential control measures. Nonpathogenic fungi (Schneider 1984, Ordentlich et al. 1991) and bacteria (Becker et al. 1990, van Peer et al. 1991, Mutitu et al. 1991) have been found to effectively reduce the incidence of infection by pathogenic, wilt-inducing Fusarium oxysporum formae speciales. Rhizobacteria stimulated celery plant growth but did not reduce disease (Becker et al. 1990). Cross-protection of the susceptible plants (usually by initial root dip with nonpathogenic spore suspension, often nonpathogenic F. oxysporum) prior to inoculation with the pathogen has also been found to significantly reduce the disease severity of Fusarium wilts (Louter and Edgington 1990, Tezuka and Makino 1991) including Fusarium yellows of celery (Schneider 1984). Brammall and Higgins (1987) suggested that the competition for infection sites between the two fungi may explain the reduction of disease severity. They also propose that the nonpathogenic fungus may stimulate host defenses, such as: (1) the accumulation of phytoalexins (Van Peer et al. 1991) in the plant which protect against subsequent infection by the pathogenic fungus, 19 (2) the manufacturing of occlusions of the primary xylem by tyloses which reduce the spread of the pathogen within the plant (MacHardy and Beckman 1981) or, (3) perhaps rapid (within 3 days) phenolic build—up within the plant (MacHardy and Beckman 1981). For over 65 years, people have recognized the existence of Fusarium-suppressive soils (T oussoun 1975). How to use this knowledge for disease control, however, is still under investigation. One of the most intensive studies of soil suppression is from the Chateaurenard region in France, where growers have been able to cultivate muskrnelon without damage caused by Fusarium oxysporum f. sp. melonis despite the presence of this pathogen in these soils (Alabouvette et al. 1979). The disease-suppressive capabilities of these soils is a result of the microbiological ecology of the soils which is influenced by the physico—chemical soil properties (Alabouvette et al. 1979, Amir and Alabouvette 1993). The action of these soils is not wholly clear, but it appears that some form of antagonism or competition amongst the soil microbes is involved. Alabouvette et al. (1979) discovered that steaming suppressive soils caused them to lose their suppressiveness, suggesting that soil microbes and not abiotic factors are responsible for the disease control. Pathogens isolated from the suppressive soils were not avirulent ecotypes (strains) since they have been found to be pathologically and morphologically similar to those in conducive soils (Nelson 1981). It has been shown for some soils that the indigenous mycoflora are necessary for suppressiveness (Alabouvette et al. 1979). In fact, it may be the populations of non-pathogenic Fusaria that antagonize or compete with the pathogens, rendering the soil suppressive (Rouxel et al. 1979), although other fungi and bacteria are important too. Opgenorth and Endo (1983) found that Corynebacterium had a suppressive effect on FOA growth, reduced the fungal activity 20 near celery roots, and reduced chlamydospore germination. In other studies, suppression was linked mainly to the inhibition of Fusarium oxysporum chlamydospore germination by the production of siderophores by fluorescent pseudomonads (Elad and Baker 1985). Soil structure and composition is fundamentally important for encouraging the appropriate microbiological composition that provides disease suppression (Alabouvette et al. 1979, Amir and Alabouvette 1993, Toussoun 1975). Compared with conducive soils, suppressive soils have greater biomass that assimilate available nutrients more rapidly (Amir and Alabouvette 1993) which suggests that indigenous microbial populations may be more aggressive competitors than pathogenic Fusaria. In suppressive soils, Fusarium oxysporum f. sp. melonis macroconidia formed germ tubes which were lysed rapidly or produced small chlamydospores. Significantly fewer chlamydospores germinated in suppressive than conducive soils in a 24 hour period (Alabouvette et al. 1979). Alabouvette et al. (1979) also found that suppressiveness could be transferred from suppressive to conducive soils even at considerable dilutions. A Michigan muck field thought to be suppressive to Fusarium yellows of celery had a slightly higher pH, higher organic content, and different nutrient levels than a nearby field conducive to the disease (Toth 1989). Although the suppressive effects may be transferrable to other soils (T oth 1989), the mechanism of suppression is unknown. Modification of the chemical environment may help to reduce disease incidence or severity. Fusarium yellows of celery have been found to increase in severity as soil pH drops from 8.3 to 3.9 (Opgenorth and Endo 1983). General Fusarium control has been provided with increasing soil pH above neutrality (Woltz and Jones 1981). The form of nitrogen available to Fusarium oxysporum f. sp. apii also appears to change the 21 severity of disease (Schneider 1985). Calcium nitrate as well as potassium chloride can help control Fusarium yellows of celery (Schneider 1985). Chitin may increase the populations of mycolytic microbes in soil (Mitchell and Alexander 1961). Amending soil with chitin has been used to control some soilbome pathogens (Mitchell and Alexander 1961) but not FOA2 (M.L. Lacy, personal communication). Organic soil amendments with crop residues have been examined for soilbome, disease control. A literature review is given in Chapter 3. LITERATURE CITED Agrios, G.N. 1988. Plant Pathology 3rd edition. Academic Press Inc. San Diego, CA. p.803. Alabouvette, C., F. Rouxel, and J. Louvet. 1979. Characteristics of Fusarium wilt- suppressive soils and prospects for their utilization in biological control. pp. 165-182. In: Soil-Borne Elan; Pathogens. eds. Schippers, B. and W. Gams. Academic Press, London. p.686. Amir, H. and C. Alabouvette. 1993. Involvement of soil abiotic factors in the mechanisms of soil suppressiveness to Fusarium wilts. Soil. Biol. Biochem. 25: 157-164. Armstrong, GM. and J.K. Armstrong. 1948. Nonsusceptible hosts as carriers of wilt Fusaria. Phytopathol. 38:808-826. Armstrong, GM. and J .K. Armstrong. 1962. Wilt of Mexican sunflower: causal Agent, the wilt of Fusarium from celery. Phytopathol. 52:723 (Abstr.) Armstrong, GM. and J.K. Armstrong. 1966. Wilt of Mexican sunflower caused by the celery-wilt Fusarium. Plant Dis. Rep. 50:391-393. Armstrong, GM. and J.K. Armstrong. 1967. The celery-wilt Fusarium causes wilt of garden pea. Plant Dis. Rep. 51:888-892. Armstrong, GM. and J .K. Armstrong. 1968. Formae speciales and races of Fusarium oxysporum causing a tracheomycosis in the syndrome of disease. Phytopathol. 58: 1242- 1246. Assigbetse, K.B., K. Fernandez, M.P. Dubois, and J.P. Geiger. 1994. Differentiation of Fusarium oxysporum f.sp. vasinfectum races on cotton by random amplified polymorphic DNA (RAPD) analysis. Phytopathol. 84:622-626. Awuah, R.T. and J.W. Lorbeer. 1986. A sorbose-based selective medium for enumerating propagules of Fusarium oxysporum f.sp. apii race 2 in organic soil. Phytopathol. 76: 1202- 1205. 22 23 Awuah, R.T. and J.W. Lorbeer. 1989. Role of light, temperature, and method of propagation in cultural variability of Fusarium oxysporum f.sp. apii race 2. Mycol. 81:278-283. Awuah, R.T. and J.W. Lorbeer. 1991. Methyl bromide and steam treatment of an organic soil for control of Fusarium yellows of celery. Plant Dis. 75: 123-125. Awuah, R.T., J.W. Lorbeer, and L.A. Ellerbrock. 1986. Occurrence of Fusarium yellows of celery caused by Fusarium oxysporum f.sp. apii race 2 in New York and its control. Plant Dis. 70:1154-1158. Banihashemi, Z. 1968. The biology and ecology of Fusarium oxysporum f.sp. melonis in soil and the root zones of host and nonhost plants. Ph.D. Dissertation. Michigan State University, East Lansing, MI. p.114. Beattie, W.R. 1907. Celery Culture. Orange Judd Co., New York. p.143. Becker, J.O., C.A. Hepfer, G.Y. Yuen, S.D. Van Gundy, M.N. Schroth, J.G. Hancock, A.R. Weinhold, and T. Bowman. 1990. Effect of rhizobacteria and metham-sodium on growth and root microflora of celery cultivars. Phytopathol. 80:206-211. Beckman, C.H. 1987. The Nature of Wilt Diseases of Plants. APS Press, St. Paul, MN. p.175. Berger, L.A. 1986. Orderly marketing: a comparative institutional analysis of coordination performance in the apple and celery subsectors. M.S. Thesis. Michigan State University, East Lansing, MI. p.217. Booth C. 1984. The Fusarium problem: historical, economic and taxonomic aspects. pp.1-13. In: The Applied Mycology of Fusarium. ed. Moss, MO. and J.E. Smith. Cambridge University Press, Cambridge. p.264. Brammall, R.A. and V.J. Higgins. 1987. The control of Fusarium crown and root rot disease of tomato by root inoculation with alien pathogenic fungi. Can. J. Plant Pathol. 9:274. (Abstr.) Burgess, L.W. 1981. General ecology of the Fusaria. pp. 225-235. In: Fusarium: Diseases, Biology, and Taxonomy. eds. Nelson, P.E., T.A. Toussoun, and R.J. Cook. Pennsylvania State University Press, University Park, Pennsylvania. p.457. Burnett, J.H. 1984. Aspects of Fusarium genetics. pp.39—69. In: The Applied Mycology pf Fusarium. ed. Moss, M.O. and J .E. Smith. Cambridge University Press, Cambridge. p.264. 24 Cerkauskas, R.F. and MR. McDonald. 1989. Race 2 of Fusarium oxysporum f.sp. apii new to Ontario. Plant Dis. 73:859. Cerkauskas, R.F. and M. Chiba. 1991. Soil densities of Fusarium oxysporum f.sp. apii race 2 in Ontario, and the association between celery cultivar resistance and photocarcinogenic furocoumarins. Can. J. Plant Pathol. 13:305-314. Cook, R.J. 1981. Water relations in the biology of Fusarium. In: Fusarium° Diseases Biology and Taxonomy. eds. Nelson, P.E., T.A. Toussoun, and R.J. Cook. Pennsylvania State University Press, University Park, Penn. p.457. Coons, GB. 1915. Stunting Disease of Celery. Mich. Agric. Exp. Stn. Annu. Rep. 28:214-215. Correll, J .C. 1992. Genetic, biochemical, and molecular techniques for the identification and detection of soilbome plant-pathogenic fungi. pp.7-16. In: Methods fpr Research 911 Soilborne Phytopathogenic Fungi. eds. Singleton, L.L., J.K. Hihail, and GM. Rush. APS Press. St. Paul, MN. p.265. Correll, J.C., C.J.R. Klittich, and LP. Leslie. 1987. Nitrate nonutilizing mutants of Fusarium oxysporum and their use in vegetative compatibility tests. Phytopathol. 77: 1640-1646. Correll, J.C., J.E. Puhalla, and R.W. Schneider. 1986a. Identification of Fusarium oxysporum f. sp. apii on the basis of colony size, virulence, and vegetative compatibility. Phytopathol. 76:396-400. Correll, J.C., J.E. Puhalla, and R.W. Schneider. 1986b. Vegetative compatibility groups among nonpathogenic root-colonizing strains of Fusarium oxysporum. Can. J. Bot. 64:2358-2361. Curl, E.A. 1988. The role of soil microfauna in plant-disease suppression. CRC Crit. Rev. Plant Sci. 72175-196. Dhingra, OD. and LB. Sinclair. 1985. Basic Plant Pathglogy Methods. CRC Press Inc., Boca Raton, FL. p.355. Dimond, A.E. and PE. Waggoner. 1953. The physiology of lycomarasmin production by Fusarium oxysporum f.sp. lycopersici. Phytopathol. 43:229-235. Edmondson, B. 1992. Selling celery. Am. Demographics 6:2. Elad, Y., and R. Baker. 1985 . The role of competition for iron and carbon in suppression of chlamydospore germination of Fusarium spp. by Pseudomonas spp. Phytopathol. 75: 1053- 1059. 25 Elmer, W.H. 1985. The ecology and control of Fusarium yellows of celery in Michigan. Ph.D. Dissertation. Michigan State University, East Lansing. p.146. Elmer, W.H. and M.L. Lacy. 1982. The reappearance of Fusarium yellows of celery in Michigan. Phytopathol. 72:1135. (Abstr.) Elmer, W.H and M.L. Lacy. 1984a. The effects of crop residues on populations of Fusarium oxysporum f.sp. apii race 2 in soil and resulting disease in celery. Phytopathol. 74:865-866 (Abstr.). Elmer, W.H. and M.L. Lacy. 1984b. Use of a color mutant of Fusarium oxysporum f.sp. apii race 2 in studies of soil population dynamics. Phytopathol. 74: 1269. (Abstr.) Elmer, W.H. and M.L. Lacy. 1987a. Effect of inoculum densities of Fusarium oxysporum f.sp. apii in organic soil on disease expression in celery. Plant Dis. 71: 1086- 1089. Elmer, W.H. and M .L. Lacy. 1987b. Effects of crop residues and colonization of plant tissues on propagule survival and soil populations of Fusarium oxysporum f.sp. apii race 2. Phytopathol. 77:381-387. Ezzell, C. 1992. Celery studies yield blood pressure boon. Sci. News 141:319. Gaye, M.M., D.J. Ormrod, F.M. Seywerd, and W.J. Odermatt. 1991. Occurrence of Fusarium yellows on celery in southwestern British Columbia and evaluation of cultivars for disease tolerance. Can. J. Plant Pathol. 13:88-92. Grajal-Martin, M.J., C]. Simon, and F .J . Muehlbauer. 1993. Use of random amplified polymorphic DNA (RAPD) to characterize race 2 of Fusarium oxysporum f.sp. pisi. Phytopathol. 83:612-614. Hart, LP. and R.M. Endo. 1975. Fusarium yellows of celery in California. Proc. Am. Phytopathol. Soc. 2:114. Hart, LP. and R.M. Endo. 1978. The reappearance of Fusarium yellows of celery in California. Plant Dis. Rep. 62:138-142. Hart, LP. and R.M. Endo. 1981. The effect of time of exposure to inoculum, plant age, root development, and root wounding on Fusarium yellows of celery. Phytopathol. 71:77-79. Hendrix, F.F. Jr. and L.W. Nielsen. 1958. Invasion and infection of crops other than the forma suscept by Fusarium oxysporum f. sp. batatas and other formae. Phytopathol. 48:224-228. 26 Huber, D.M. and RD. Watson. 1970. Effect of organic amendment on soil-borne plant pathogens. Phytopathol. 60:22—26. Iannelli, D., R. Capparelli, F. Marziano, F. Scala, and C. Noviello. 1983. Production of hybridomas secreting monoclonal antibodies to the genus Fusarium. Mycotaxon 17:523-532. Ireland, K.F. and M.L. Lacy. 1986. Laboratory identification of pathogenic isolates of Fusarium oxysporum f.sp. apii race 2. Phytopathol. 76:956. (Abstr.) Jordan, C. M., R.M. Endo, and LS. Jordan. 1988. Penetration and colonization of resistant and susceptible Apium graveolens by Fusarium oxysporum f.sp. apii race 2: callose as a structural response. Can. J. Bot. 66:2385-2391. Jordan, C.M., L.S. Jordan, and R.M. Endo. 1989. Association of phenol-containing structures with Apium graveolens and resistance to Fusarium oxysporum f.sp. apii race 2. Can. J. Bot. 67:3153-3163. Katan, J. 1971. Symptomless carriers of the tomato Fusarium wilt pathogen. Phytopathol. 61:1213-1217. Kistler, H.C, P.W. Bosland, U. Benny, S. Leong, and P.H. Williams. 1987. Relatedness of strains of Fusarium oxysporum from crucifers measured by examination of mitochondrial and ribosomal DNA. Phytopathol. 77:1289-1293. Kistler, H.C., E.A. Momol, and U. Benny. 1991. Repetitive genomic sequences for determining relatedness among strains of Fusarium oxysporum. Phytopathol. 81 :331-336. Lacy, M.L. and E. Grafius. 1980. Diseases and insect pests of celery. Michigan State University Ext. Bull. E-1427. p.8. Lorenz, O.A. and D.N. Maynard. 1988. Knott’s Handbgzk for Vegetable Qreyyers - Third flitien. John Wiley & Sons, Inc., New York. p.456. Louter, J .H. and L.V. Edgington. 1990. Indications of cross-protection against Fusarium crown and root rot of tomato. Can. J. Plant Pathol. 12:283-288. MacHardy, W.E. and C.H. Beckman. 1981. Vascular wilt Fusaria: infection and pathogenesis. pp. 365-390. In: Fusarium: Diseases, Biology, and Taxonomy. eds. Nelson, P.E., T.A. Toussoun, and R.J. Cook. Pennsylvania State University Press, University Park, Pennsylvania. p.457. Manicom, B.Q., M. Bar-Joseph, J.M. Kotze, and M.M. Becker. 1990. A restriction fragment length polymorphism probe relating vegetative compatibility groups and pathogenicity in Fusarium oxysporum f.sp. dianthi. Phytopathol. 80:336-339. 27 Manulis, S., N. Kogan, M. Reuven, and Y. Ben-Yephet. 1994. Use of RAPD technique for identification of Fusarium oxysporum f.sp. dianthi from carnation. Phytopathol. 84:98-101. Marchant, R. 1984. The ultrastructure and physiology of sporulation in Fusarium. pp.15-37. In: The Applied Mycology of Fusarium. ed. Moss, M.O. and J.E. Smith. Cambridge University Press, Cambridge. p.264. Martyn, R.D., L.W. Barnes, and J. Amador. 1987. Fusarium yellows of celery in Texas. Plant Dis. 71:651. Michigan Department of Agriculture. 1993. Michigan Agricultural Statistics. Mitchell, R. and M. Alexander. 1961. The mycolytic phenomenon and biological control of Fusarium in soil. Nature 190:109-110. Mutitu, E.W., D.M. Mukunya, and 8.0. Keya. 1991. Control of Fusarium yellows of beans using bacterial antagonists. Discovery Innovation 3:83-86. (Abstr.) Nelson, P.E., T.A. Toussoun, and R.J. Cook. 1981. Fasgium; Disc-fies, Bielegy, ape fIfaxenemy. The Pennsylvania State University Press, University Park, Pennsylvania. p.457. Nelson, P.E., T.A. Toussoun, and W.F.O. Marasas. 1983. Fusarium Species: An Illustrated Maneal for Identification. The Pennsylvania State University Press, University Park, Pennsylvania. p.193. Nelson, R., 6.11. Coons, and LC. Cochran. 1937. The Fusarium yellows disease of celery (Apium graveolens L. var. dulce D.C.). Mich. State Univ. Agric. Exp. Stn. Tech. Bull. No. 155. p.74. Newhall, A.G. 1953. Blights and other ills of celery. pp.408-417. In: Plant Disgses. USDA Yearbook of Agriculture. Washington, D.C. p.940. Opgenorth, D.C. and R.M. Endo. 1979. Sources of resistance to Fusarium yellows of celery in California. Plant Dis. Rep. 63:165-169. Opgenorth, D.C. and R.M. Endo. 1981. Competitive saprophytic ability (CSA) of Fusarium oxysporum f.sp. apii. Phytopathol. 71:246-247. Opgenorth, D.C. and R.M. Endo. 1983. Evidence that antagonistic bacteria suppress Fusarium wilt of celery in neutral and alkaline soils. Phytopathol. 73:703-708. Opgenorth, D.C. and R.M. Endo. 1985. Additional sources of resistance to race 2 of Fusarium oxysporum f.sp. apii. Plant Dis. 69:882-884. 28 Ordentlich, A., Q. Migheli, and I. Chet. 1991. Biological control activity of three Trichoderma isolates against Fusarium wilts of cotton and muskmelon and lack of correlation with their lytic enzymes. J. Phytopath. 133:177-186. Orton, T.J. 1981. Physical association of Fusarium oxysporum f.sp. apii with germinating celery seeds following artificial inoculation. PhytOpath. Z. 102:320-327. Otto, H.W., A.O. Paulus, M.J. Snyder, R.M. Endo, L.P. Hart, and J. Nelson. 1976. A crown rot of celery. Cal. Agr. 30:10-11. Papavizas, G.C. 1975. Crop residues and amendments in relation to survival and control of root-infecting fungi: an introduction. p.76. In: Biology and Contrel of Soil- Borne Plant Pathogens. ed. Bruehl, G.W. APS Press, St. Paul, MN. p.216. Park, D. 1954. Chlamydospores and survival in soil fungi. Nature 173:454-455. Park, D. 1959. Some aspects of the biology of Fusarium oxysporum Schl. in soil. Ann. Bot. 23:35-49. Price, D. 1984. Fusarium and plant pathology: the reservoir of infection. pp.71-93. In: The Applied Mycology of Fusarium. eds. Moss, M.O. and J.E. Smith. Cambridge University Press, Cambridge. p. 264. Puhalla, J .E. 1981. Genetic considerations of the genus Fusarium. pp.291-305. In: Easasium: Diseases, Biolegy and Taxonomy. eds. Nelson, P.E., T.A. Toussoun, and R.J. Cook. Pennsylvania State University Press, University Park, Penn. p.457. Puhalla, J .E. 1984a. A visual indicator of heterokaryosis in Fusarium oxysporum from celery. Can. J. Bot. 62:540-545. Puhalla, J .E. 1984b. Races of Fusarium oxysporum f.sp. apii in California and their genetic interrelationships. Can. J. Bot. 62:546-550. Puhalla, J.E. 1985 . Classification of strains of Fusarium oxysporum on the basis of vegetative compatibility. Can. J. Bot. 63:179-183. Ralton, J .E., M.G. Smart, and AB. Clarke. 1987. Recognition and infection processes in plant pathogen interactions. pp. 217-252. In: l_’l_ant-Microbe Interactions: Molecular and Genetic Perspectives Vol.2. eds. Kosuge T. and E.W. Nester. Macmillan Pub. Co., New York. p.448. Rouxel, F ., C. Alabouvette, and J. Louvet. 1979. Recherches sur la resistance des sols aux maladies. IV - Mise en evidence du role des Fusarium autochtones dans la resistance d’un sol a la Fusariose vasculaire du melon. Ann. Phytopath. 11:199-207. 29 Ryder, E]. 1979. Leafy Salad Vegetables. AVI Publishing Co., Westport. p.266. Ryker, T.C. 1935. Fusarium yellows of celery. Phytopathol. 25:578-599. Salttink, OJ. and A.E. Dimond. 1964. Nature of plugging material in xylem and its relation to rate of water flow in Fusarium-infected tomato stems. Phytopathol. 54:1137- 1140. Scheffer, R.P. and J.C. Walker. 1953. The physiology of Fusarium wilt of tomato. Phytopathol. 43:116-125. Schneider, R.W. 1984. Effects of nonpathogenic strains of Fusarium oxysporum on celery root infection by F usarium oxysporum f. sp. apii and a novel use of the lineweaver- burk double reciprocal plot technique. Phytopathol. 74:646-653. Schneider, R.W. 1985. Suppression of Fusarium yellows of celery with potassium, chloride, and nitrate. Phytopathol. 75:40-48. Schneider, R.W. and J .L. Norelli. 1981. A new race of Fusarium oxysporum f.sp. apii in California. Phytopathol. 71:108. (Abstr.) Snowdon, A.L. 1992. Color Atlas of Post-Harvest Diseases and Disorders ef Fppits and Vegetables Vol 2. CRC Press. Boca Raton, FL. p.416. Snyder, W.C. and H.N. Hansen. 1940. The species concept in Fusarium. Am. J. Bot. 27:64-67. Sobti, R.C., O.P. Mittal, A. Sachdeva, and GB. Gill. 1991. Anticlastogenic effect of essential oil of seeds of Apium graveolens. Cytol. 56:303-308. Tezuka, N. and T. Makino. 1991. Biological control of Fusarium wilt of strawberry by nonpathogenic Fusarium oxysporum isolated from strawberry. Ann. Phytopathol. Soc. Jpn. 57:506-511. Toth, K.F. 1989. Biology and control of Fusarium oxysporum f.sp. apii race 2. Ph.D. Dissertation. Michigan State University, East Lansing, MI. p.174. Toth, K.F. and M.L. Lacy. 1990. Selective medium for nitrate nonutilizing mutants of Fusarium oxysporum. Phytopathol. 80:1045. (Abstr.) Toth, K.F. and M.L. Lacy. 1991. Comparing vegetative compatibility and protein banding patterns for identification of Fusarium oxysporum f.sp. apii race 2. Can. J. Microbiol. 37:669-674. 30 Toussoun, T.A. 1975. Fusarium-suppressive soils. pp.145-151. In: Biology and ' Control of Soil-Borne Plant Pathogens. ed. Bruehl, G.W. APS Press, St. Paul, MN. p.216. United States Department of Agriculture. 1993. Agricultural Statistics 1992. US Government Printing Office, Washington, D.C. p.524. Van Peer, R., G. J. Niemann, and B. Schippers. 1991. Induced resistance and phytoalexin accumulation in biological control of Fusarium wilt of carnation by Pseudomonas sp. strain WCS417r. Phytopathol. 81: 728- 734. Waite, B.H. and V.C. Dunlap. 1953. Preliminary host range studies with Fusarium oxysporum f.sp. cubense. Plant Dis. Rep. 37:79-80. Windels, CE. 1992. Fusarium. pp.115-128. In: Methods for Research on Soilborne Phytopathogenie Fungi. eds. Singleton, L.L., J .K. Hihail, and GM. Rush. APS Press. St. Paul, MN. p.265. Woltz, SS. and J .P. Jones. 1981. Nutritional requirements of Fusarium oxysporum: basis for a disease control system. pp. 340-349. In: Fusarium; Diseases, Biology, and Taxonomy. eds. Nelson, P.E., T.A. Toussoun, and R.J. Cook. Pennsylvania State University Press, University Park, Pennsylvania. p.457. Wong, W.C., M. White, and LG. Wright. 1988. Production of monoclonal antibodies to Fusarium oxysporum f.sp. cubense race 4. Lett. Appl. Microbiol. 6:39-42. Zakaria, M.A. 1978. Reduction in Fusarium populations in soil by oilseed meal amendments. PhD. Dissertation. Michigan State University, East Lansing, MI. p.86. Zheng, G., P.M. Kenney, J. Zhang, and L.K.T. Lam. 1993. Chemoprevention of benzo[a]pyrene-induced forestomach cancer in mice by natural phthalides from celery seed oil. Nutr. Cancer 19:77-86. CHAPTER 2 - SOMACLONAL VARIATION TO IMPROVE RESISTANCE OF CELERY TO FUSARIUM YELLOWS INTRODUCTION Host resistance is considered to be the most effective practical control for vascular wilt diseases (Opgenorth and Endo 1985). Control of Fusarium yellows of celery with chemicals (M.L. Lacy personal communication, Otto et al. 1976), soil amendments (Elmer 1985, Toth 1989, Schneider 1985), and root microorganism preinoculations (Becker et a1. 1990, Schneider 1984) have either not been successful or economically feasible. Host resistance used with crop rotation has been suggested to be the most effective method for controlling Fusarium yellows of celery (Toth 1989). Conventional breeding for disease resistance involves hybridizing two dissimilar cultivars, varieties, species or genera in the attempt to select for desirable features of each. Following successful hybridization, the new hybrid is backcrossed, plants are selected for desirable traits, and backcrossing and selection are continued for several generations with variety release occurring by year 7-8 at the earliest (Evans et al. 1984). Variety development may take even longer if subsequent hybridizations are done to improve other characteristics (Honma et a1. 1986). Also, varieties bred for one trait may not retain all other desirable characteristics (Orton et al. 1984b). Conventional breeding of celery lines to develop resistance to Fusarium yellows of celery has been moderately 31 32 successful. Initially, available cultivars were screened to determine existing levels of disease resistance. Most Apium graveolens var. dulce (celery) cultivars were found to be susceptible to Fusarium yellows (Elmer and Lacy 1984) but some natural variation existed. Tall Utah 52-70 HK and Deacon had moderate field resistance in Michigan in early trials (Elmer and Lacy 1984). Tall Utah 52-70 HK, Tendercrisp, and Golden Detroit showed the best available field resistance in California (Opgenorth and Endo 1985, Otto et a1. 1976), and Tall Utah 52-70 HK, Tendercrisp, Deacon, Bishop and Fordhook were moderately resistant to the disease in New York (Awuah et al. 1986). Of other Apium spp. tested, celeriac (Apium graveolens var. rapaceum) and smallage (Apium graveolens var. secalinum) were found to be moderately resistant and susceptible, respectively to Fusarium yellows (Opgenorth and Endo 1979, 1985, Awuah et al. 1986). These screenings indicated that celery crosses with celeriac might effectively control the disease and that moderately resistant cultivars might be able to be grown in fields with relatively low disease pressure. Orton et al. (1984b) developed a Fusarium yellows-resistant celery line, UC-l, that used celeriac as the source for resistance. While this line may be useful as germplasm for further cultivar development, the line itself was horticulturally unacceptable for the fresh market (Orton et al. 1984b). Celeriac was also in the pedigree for the celery cultivars Pilgrim and Companion, which both expressed moderate Fusarium yellows resistance (Honma et a1. 1986). Parsley (Petroselinum crispum (Mill.)), another genus in the Apiaceae family, is highly resistant to Fusarium oxysporum f. sp. apii infection (Elmer et al. 1986) and was once hybridized with celery (Honma and Lacy 1980). This intergeneric cross yielded only three individual plants that were the result of crosses rather than of self-pollination, using the 33 recessive yellow color trait in the female parent as a genetic marker. This cross between celery and parsley resulted in moderate resistance but poor horticultural characteristics in offspring (Honma and Lacy 1980). Backcrosses to improve horticultural traits resulted in lower levels of disease resistance. Tissue culture was at first expected to be useful for the asexual (clonal) propagation of plants, but phenotypic variants were so prevalent that clones were not as easy to regenerate as had been hoped (Larkin and Scowcroft 1981, Browers and Orton 1982a). These "artifacts" (Larkin and Scowcroft 1981) were initially discarded until it was realized that they could be used as a genetic source for crop improvement. Tissue culture turned out to be an "unexpectedly rich and novel source of genetic variability“ (Larkin and Scrowcroft 1981). Cell and tissue culture examination has shown that genetic variability is very common in plant cell cultures (Orton 1983). This genetic variability, which is often phenotypically expressed in regenerated plants ("somaclones"), has been termed "somaclonal variation" (Larkin and Scrowcroft 1981). The amount of somaclonal variation depends upon 1) plant growth regulator type and concentration used in callus initiation media (Kallak and Vapper 1985), 2) length of time in culture (Heath- Pagliuso et a1. 1988), 3) source of explant tissue (Browers and Orton 1982a), 4) speed of callus growth (Orton 1984), and 5) ploidy level of the parent line (Orton 1984). Developing disease resistant crops using tissue and cell culture techniques may provide novel sources of resistance and may take half the time that conventional breeding takes (Evans et a1. 1984) with similar levels of success. The first in vitro attempts to screen for FOA2-resistant germplasm was done by exposing celery seed to FOA2 spore-mycelial suspensions (Orton 1982). This technique 34 was an unsatisfactory method for identifying celery germplasm with Fusarium yellows resistance and would be better used to identify the virulence of FOA2 isolates (Orton 1982). Disease-resistant somaclones have been developed faster and more reliably by using parent material with moderate resistance (Wright and Lacy 1988), and by adding a pathogen-produced host toxin or culture filtrates to the selection medium (Daub 1986, Hartman et al. 1984). Fusaric acid, a Fusarium-produced toxin, did not enhance the generation of celery somaclones resistant to Fusarium yellows, suggesting that FOA2 does not require fusaric acid for disease expression (Heath-Pagliuso et al. 1988). Two toxins produced in FOA2-infected celery are known to prevent the growth of celery cells susceptible to FOA2 but they are not toxic to the resistant celery cells (Banks-Izen et al. 1981). Once the compounds are identified, these host-produced toxins could be used for in vitro selection of FOA2-resistant celery cells. Brettell and Ingram (1979) explained that disease resistance is difficult to select for in vitro because single cells or callus may not express resistance, and it is difficult in practice to expose cells to inoculum and reisolate resistant cells for plantlet regeneration. Prescreening regenerated plants on mycelial mats of FOA2 prior to screening in infested soil did help to select for more resistant plants (Pulhnan and Rappaport 1983), but no prescreened somaclones survived to maturity (Heath-Pagliuso et al. 1988). Somaclones screened in California for Fusarium yellows resistance using artificially infested soil did not produce any plants that were resistant in early trials (Pullman et al. 1984). During later attempts, one celery somaclone was produced from the highly susceptible cultivar Tall Utah 52-70 R (Heath-Pagliuso et al. 1989). The line developed from this plant, UC-T3, has been recommended for breeding with existent 35 cultivars to provide disease resistance (Heath-Pagliuso and Rappaport 1990). In Michigan, celery somaclones expressing resistance to Fusarium yellows were screened in infested soil, without prescreening in vitro (Ireland et a1. 1987, Toth and Lacy 1991). Celery somaclones with improved resistance have been regenerated from moderately resistant cultivar Tall Utah 52-70 HK (Wright and Lacy 1988, Toth and Lacy 1991), and from the highly susceptible cultivars Tall Utah 52-70 R (Heath-Pagliuso et a1. 1989) and Florida 683 (Wright and Lacy 1988). Somaclonal variation enables selection from regenerated plants for a desirable trait. This variation, it appears, encompasses both preexisting variability and induced genomic alterations/ mutations (Evans et a1. 1984). Culture-induced genetic changes have been attributed to: gene mutation, gene amplification and deletion, mitotic crossing over, karyotype alterations, chromosome rearrangement, transposable or controlling elements, cryptic viral elimination, and organelle mutation (Evans et al. 1984, Larkin and Scowfield 1981). The importance of these effects for the development of somaclones from many different crops has been reviewed extensively (Brettel and Ingram 1979, Daub 1986, Evans et a1. 1984, Larkin and Scowcroft 1981, Orton 1984). By examining segregation data, resistance in celery to Fusarium oxysporum f .sp. apii race 2 not derived from somaclonal variation has been found to be controlled by a single major dominant gene and other independent quantitative genes (Orton et al. 1984a). Different celery cultivars may have different genes for resistance (Orton et al. 1984a). Potentially, more resistance genes could be brought into a new cultivar through breeding two cultivars known to have different genes for resistance. Heath-Pagliuso and Rappaport (1990) determined that there are two loci that control the disease resistance expressed by the 36 celery somaclone UC-T3 developed in California. The genes controlling resistance of the Michigan somaclones (Toth and Lacy 1991) are not known. Although there is evidence of genetic changes in vivo (Orton 1984), the frequency of these gene and chromosomal alterations have not been determined (Orton 1984). Cultured celery cells and regenerated plants often exhibit gross chromosomal abnormalities that are not found in cells from plants that have never been subjected to tissue culture (Browers and Orton 1982a, Murata and Orton 1983). Also, the frequency of somaclonal variation is too high to be accounted for by point mutations that may arise in viva (Larkin and Scowfield 1981). These points provide indirect evidence that somaclonal variation is, in part, produced by tissue culturing and is not solely an expression of naturally occurring variation in the parent plant. Chromosomal examination of cultured cells has supported this idea. Unlike cells from normal tissue, cells from celery callus have ploidy level alterations, chromatin bridges, lagging chromosomes, multipolar spindles, centromere relocation, and changes in visible chromosome constrictions (Browers and Orton 1982a, Murata and Orton 1983). Celery callus from which cells were regenerated into plants mostly contained cells that were diploid (80%) (Browers and Orton 1982b). The non-diploids were hypoploid (11%), polyploid (3%), or had an intermediate ploidy level (6%). Some of the difficulty in assessing the source of genetic variation seen in somaclones is that evidence of chromosomal alterations in cultured cells does not mean that viable somaclones can be regenerated from these abnormal cells. Browers and Orton (1982b) showed that changes in chromosome numbers often prevent regeneration, although hypodiploids were occasionally produced. Hypodiploids may be regenerated because the apparent loss of 37 chromosome(s) may not result in a loss of genetic information. In tissue culture, Robertson fusions are often produced, during which acrocentric or telocentric chromosomes fuse to produce a metacentric chromosome (Browers and Orton 1982b, Murata and Orton 1984). Soon after regeneration, these hypodiploids may function normally, but over time they may show disruptions in cellular functions from delayed chromatid separation of the fused chromosome (Murata and Orton 1984). Inaccurate timing of chromatid separation would produce daughter cells with different karyotypes. In cases where chromosome loss was reported (Murata and Orton 1983), this effect was not a random event. Certain chromosomes appear to have characteristics that enable their disappearance without apparent damage to cellular function; these chromosomes probably have information that is not essential for growth in vitro (Murata and Orton 1983). Cells that lose a chromosome, however, may not regenerate viable somaclones. Also, diploids from cultured cells are not necessarily normal. Fifty-six percent of diploid celery somaclones were found to have chromosomes with segments of DNA missing (Murata and Orton 1983). Comparisons of chromosomal morphologies have been made between precallus and cultured celery tissue (Murata and Orton 1983). Cultured cells have more variation than uncultured cells in the amount of chromosomal constrictions and secondary constrictions that had not existed previously (Murata and Orton 1983). Tissue culturing also induced changes in centromere locations. Murata and Orton (1983) found that, on average, tissue culturing of celery induced a loss of 2-3 acrocentric chromosomes. Celery somaclones with improved resistance to Fusarium yellows were produced from Tall Utah 52-70 HK and Florida 683 parent cultivars (Wright and Lacy 1988). 38 Moderately resistant Tall Utah 52-70 HK-derived somaclones were used at first to develop resistant cultivars because more disease resistant plants were regenerated and the observed resistance was more likely to be stable (Toth and Lacy 1991). By 1992, these somaclones had been selected for two (R2) or three (R3) generations, depending upon the line (Krebs and Grumet unpublished). Successive screenings and selections have shown disease resistance in these somaclones to be heritable and capable of improvement (Krebs and Grumet unpublished, Toth and Lacy 1991). The successful development of TU 52- 70 I-IK somaclones with improved resistance prompted efforts to produce Florida 683- derived somaclones (T oth and Lacy unpublished). Florida 683 is a celery cultivar valued by Michigan growers for its compactness, fuller heart and higher weight. A Florida 683 cultivar with improved resistance would be the cultivar of choice for Michigan growers (T. Dudek, personal communication). Selections of the first generation of Florida 683- derived somaclones had been completed at the onset of my project (Krebs and Grumet unpublished). ' The objective of this research was to continue selection of the Tall Utah 52-70 HK—derived somaclones and the Florida 683-derived somaclones for their disease resistance and horticultural desirability. 39 MATERIALS AND METHODS Somaclone Planting Celery somaclone seed was treated with tetramethyl-thiuram disulfide (Thiram9), planted in Baccto" professional potting mix (Michigan Peat Company) in plug trays, and kept at 24 j; 4°C with a 16 hour photoperiod. Upon emergence, seedling trays were fertilized bi-weekly with 20-30 ml each of a 65 ppm (4.9 g/l) Peters‘D 20-20-20 (N -P-K) soluble fertilizer. At 7 weeks, 30 seedlings per line were planted in 4.5 111 rows in fields naturally infested with Fusarium oxysporum f. sp. apii race 2. Replicated plots were planted in four randomized blocks and unreplicated lines were planted in single plots in randomly chosen locations. In 1992, R1 and R2 Florida 683-derived somaclones and R., Tall Utah 52-70 HK- derived somaclone lines derived from single plants were planted in unreplicated plots at Willbrandt Farms, Muskegon, MI. Seeds from all related lines derived from a single parent were combined for replicated trials. These were called bulked lines. Four replications of R4, bulked Tall Utah 52-70 HK somaclones were also set up in a randomized complete block design for field evaluation at this farm. Florida 683, Tall Utah 52-70 HK, and Picador were used as susceptible, moderately resistant and highly resistant controls, respectively. In 1993, R2 and R3 FL 683 somaclones and R., and R5 TU 52-70 HK somaclones were planted in unreplicated plots at Wilbrandt Farms, Decatur, MI. Replicated variety trials of the bulked TU 52-70 HK somaclones of the R5 generation (F128(3)-1, F128(4), K26-1, K128, K108(3)-2) were planted at VerHage Farms, Hudsonville, MI. Florida 40 683, Tall Utah 52-70 HK, and Picador were used as susceptible, moderately resistant and highly resistant controls at both field sites. In 1994, two plantings of R., bulked FL 683 somaclones were made at the Jim Holstege Farm, Grand Rapids, MI and at Bruce Von Solkema Farm, Byron Center, MI. Each planting was designed as a randomized complete block with 3 replications. At all locations, Florida 683, Tall Utah 52-70 HK, and Picador were used as replicated susceptible, moderately resistant, and highly resistant controls, respectively. Unreplicated plots of all individual FL 683 somaclonal lines (R3 and R.,) were also planted at the Holstege and Van Solkema Farms. Florida 683 and Picador were used as susceptible and resistant controls. Variety trials of R5 bulked HK somaclones were planted in a randomized complete block design with four replicates at the VerHage Farm, Hudsonville, MI. These lines, designated SHKl, SHK2, SHK3, SHK4, and SHKS, were previously called F128(3)-1, F128(4), K26-1, K108(3)-2, and K128, respectively. The SHK (meaning somaclone HK-derived) seed had been multiplied in the field by planting isolated plots composed of the half-sib families of each line. Within the isolated plots, pollination was not controlled. The seed increase was done by Petoseed Company, Arroys Grande, California. Harvesting and Evaluation At harvest, ten plants from each line and ten from each replication of the bulk lines were chosen randomly. Plants were cut diagonally through the crowns and rated for Fusarium yellows disease on a scale of 1-5 (1 = no vascular discoloration, 2 = trace of vascular discoloration, 3 = less than 50% crown area discolored but more than a 41 trace, 4 = 50% or more crown area discolored, and 5 = dead or nearly dead plant). Plants were also rated for their horticultural acceptability. In 1992 and 1993, the characteristics measured were compactness (1 = most compact; 3 = least compact), twisting (1 = no twisting; 3 = extreme twisting), and pithiness (1 = not pithy; 3 = very pithy). In 1993, the amount of suckering was also rated (1 = few suckers; 3 = many suckers). In 1994, plants were rated for stalk shape appearance (compactness, petiole twisting), external markings (growth cracks, discoloration), and internal acceptability (hollow or pithy petioles). Each characteristic was rated on a 1 to 3 scale, where 1 = most desirable and 3 = least desirable. Each plant was trimmed to 35 cm long and outer malformed petioles and suckers were removed. The ten trimmed plants from each group were weighed and this value was converted to marketable yield in crates per acre using: crates/acre = ((lb/plot)/(5.4/(43560/2.67)))/55, where: 55 = lb/crate 5.4 = number of ft/row length 2.67 = distance between rows in feet 43560 = number of ftZ/acre. The statistical packages PC-SAS (SAS Institute Inc., Carry, NC) and SigmaStat (Jandel Scientific Software, San Rafael, CA) were used to compare all of the lines using a randomized complete block design for the replicated plots (P=0.05). Tukey’s test (P=0.05) was used for mean separation. Unreplicated plots were not statistically analyzed. To make comparisons among different locations and years, average relative disease ratings were calculated by dividing the average disease rating of the somaclones by the average disease rating of the parent cultivar. 42 The most disease-free plants were selected for seed production. Guidelines for selection were as follows: a) no plants were kept from lines with more than two of ten plants with a disease rating > 2; b) yields of 2 7.2 kg per ten plants were acceptable but 2 8.5 kg was desired; c) most horticulture characteristics rated as 1 but some 2 ratings were still acceptable if the disease ratings were good. Plants from these selected lines were uprooted and 1/3 of the crown was removed diagonally down through the main root to ensure that plants were symptom-free. Between 3 and 16 plants per line exhibiting no vascular discoloration were trimmed to 35 cm (most of the leaves were removed to reduce transpiration rate) and taken to the greenhouse for planting. Preparation of Plants for Vernalization In the greenhouse, crowns were washed and then sprinkled with RooTone" powder (Pratt-Gabriel, Hanover, PN) and the plants were transplanted into Baccto" professional potting mix (Michigan Peat Company, Houston, TX) in 30 cm pots. These pots were set under greenhouse benches for one week to protect from direct sunlight and to lower water demand. In the second week, plants were covered with cheesecloth and placed on greenhouse benches. The celery plants were watered lightly and kept at 24 j; 2°C under a 14 hour photoperiod for about four weeks prior to vernalization. It was important not to overwater at this stage to prevent rot. Twice a week, a 0.1 M calcium chloride solution was sprayed on the apical meristem to prevent black heart. Plants were fertilized weekly with either 100 ml full strength Hoagland’s solution or 50 ml of 65 ppm soluble 20-20-20 (N-P-K), alternating weekly. 43 Vernalization Four to five weeks after harvest, celery plants were placed in the vemalization chamber at the Horticulture Farm, MSU campus and maintained at 4°C with a 12 hour photoperiod (four-100 Watt fluorescent bulbs). Celery plants were sprayed with a 0.1 M calcium chloride solution 2 or 3 times per week and watered twice weekly for 3 weeks and then once a week thereafter. Once monthly, 100 ml of full strength Hoagland’s solution was given to each plant. Greenhouse Conditions Post-Vemalization and Seed Collection After 63-70 days in vemalization, plants were placed in the greenhouse at 21 j; 2°C with a 16 hour photoperiod. Plants were watered as needed and treated with 0.1 M calcium chloride three times per week until bolting began. Each plant was given 100 ml full strength Hoagland’s solution and 50 ml of 65 ppm 20-20-20 (N -P-K) soluble fertilizer in alternate weeks. As flowering began, plants were enclosed in cheesecloth cages and shaken to self-fertilize. Twelve weeks post-vemalization, fertilizer treatments were stopped and water was withheld to encourage seed production and drying. Mature celery seed from each plant was collected by hand as seed turned brown and dry. Seed was then air dried at room temperature (22°C) to prevent mold. Dried seed was cleaned of debris using a small forced-air seed cleaner, weighed, and stored at 4°C. 44 RESULTS Tall Utah 52-70 HK-derived somaclone screening In 1992, all TU 52-70 HK—derived somaclonal bulked lines from the R4 generation had lower disease ratings than the moderately resistant parent cultivar (Table 1), Tall Utah 52-70 HK, but the ratings were not statistically significant at P=0.05. These somaclonal lines and TU 52 70 HK also had disease ratings statistically comparable to the Fusarium yellows-resistant cultivar Picador. All lines had significantly lower disease ratings than the susceptible control, Florida 683. Yields from all lines were statistically equivalent (Table 1). The individual somaclonal families used to form the bulked lines were also field screened to detect any individuals with poor performance. Most TU 52- 70 HK-derived somaclones had lower disease ratings than TU 52—70 HK, however K26- 1-7 and K128-5 had the same average disease ratings as TU 52-70 HK (Table 2). Yields ranged from 764-1181 crates/acre; TU 52-70 HK produced 1077 crates/acre. Horticultural traits of plant compactness, petiole twisting, and petiole pithiness scored well in 1992 for most TU 52-70 HK-derived somaclones (Table 2). Somaclonal lines K128-13, K26-1-4, and K26-1-5 had the worst overall appearance. Somaclones that rated worst for each individual characteristic were from the K128 line: K128-17 was least compact, K128-13 has the most petiole twisting, and K128-16 was the most pithy (Table 2). In 1993, bulked somaclonal lines from the R5 generation had lower disease ratings than the parent cultivar, TU 52-70 HK, although only the K108(3)-2 line was significantly lower at P=0.05 (Table 3). Yields from all lines except Florida 683 were 45 Table 1. Reactions of bulk Tall Utah 52-70 HK—derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in a replicated field trial near Muskegon, MI, 1992. Yield Celery Line Disease Rating1 (55 lb crates/acre) F128(3) (R.) 1.2 b2 928 a2 F128(4) (R.) 1.1 b 994 a K26-1 (R.,) 1.1 b 972 a K128 (R.,) 1.3 b 983 a K108(3)-2 (R.,) 1.1 b 989 a Controls: Florida 683 2.1 a 1000 a Tall Utah 52-70 HK 1.5 b 1077 a Picador 1.1 b 1066 a 1 Disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Within columns, means followed by the same letter were not significantly different according to Tukey’s Test (P = 0.05). 46 Table 2. Reactions of R4 Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in an unreplicated field trial near Muskegon, MI, 1992. ‘ Disease Yield (55 lb 1 Celery Line Ratingl crates/ acre) Compact2 Twisting2 Pithy2 Fl28(3)1-1 1.1 901 1.0 1.0 1.0 Fl28(3)1-2 1.0 906 1.0 1.0 1.0 F128(3)l-4 1.0 917 1.0 1.0 1.0 Fl28(3)1-5 1.4 851 1.0 1.0 1.0 F128(4)1 1.1 961 1.0 1.0 1.0 F128(4)2 1.3 1099 1.2 1.0 1.0 F128(4)3 1.1 989 1.1 1.2 1.0 F128(4)4 1.2 912 1.0 1.0 1.0 F128(4)5 1.1 1055 1.1 1.0 1.0 Fl28(4)6 1.0 1077 1.0 1.3 1.0 F128(4)7 1.1 945 1.0 1.1 1.0 F128(4)“ 1.1 945 1.0 1.0 1.0 K26-1-1 1.0 917 1.3 1.0 1.0 K26-1-2 1.0 775 1.0 1.0 1.0 K26-1-3 1.4 967 1.3 1.0 1.1 K26-l-4 1.1 1099 1.4 1.1 1.3 K26-1-5 1.2 967 1.4 1.0 1.3 K26-l-6 1.1 824 1.2 1.1 1.3 K26-1-7 1.5 780 1.0 1.1 1.0 K108(3)2-l 1.1 967 1.0 1.0 1.0 K108(3)2-3 1.2 967 1.5 1.0 1.0 K108(3)2-4 1.1 769 1.0 1.0 1.0 K108(3)2-5 1.2 802 1.0 1.0 1.0 K128-1 1.2 1000 1.2 1.0 1.0 K128-2 1.4 928 1.4 1.2 1.0 K128-3 1.4 764 1.0 1.0 1.0 K128-4 1.1 994 1.2 1.0 1.0 K128-5 1.5 1181 1.2 1.0 1.0 K128-7 1.4 1071 1.3 1.0 1.2 K128-13 1.4 901 1.5 1.5 1.3 K128-15 1.3 890 1.0 1.0 1.0 K128-l6 1.1 1016 1.0 1.0 1.4 K128-l7 1.4 879 1.6 1.0 1.0 K128-l8 1.4 868 1.2 1.0 1.1 Control: TU52-70HK l .5 1077 - - - ' Disease ratings were based on a l to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Mean value of ten plants rated for compactness, twisting, and pithiness using a rating scale of l to 3, where 1 =most desirable and 3 =least desirable. 47 Table 3. Reactions of R, bulk Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in a replicated field trial near Hudsonville, MI, 1993. Disease Yield (55 lb Celery Line Ratingl crates/acre) Compact2 Twistingz Suckers2 Pithy2 K108(3)-2 1.3 a3 796 a3 1.3 1.0 1.7 1.0 F128(3)-l 1.5 ab 807 a 1.1 1.1 1.4 1.1 F128(4) 1.5 ab 818 a 1.2 1.1 1.2 1.1 K26—1 1.4 ab 829 a 1.2 1.2 1.3 1.2 K128 1.4 ab 906a 1.3 1.2 1.4 1.1 Controls: Picador 1.2 a 928 a 1.0 1.0 1.8 1.1 TU52-70HK 1.9 b 719 a 1.6 1.7 1.4 1.1 Florida 683 4.0 c 27 b - - - - ‘ Disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Mean value of ten plants rated for compactness, twisting, suckering, and pithiness using a rating scale of 1 to 3, where 1 = most desirable and 3 = least desirable. 3 Within columns, means followed by the same letter were not significantly different according to Tukey’s Test (P = 0.05). 48 statistically equivalent. TU 52-70 HK-derived somaclones screened at Decatur, MI had lower disease (Table 4) than those screened at Hudsonville, MI (Table 3) which reflected differences in the disease pressure at each site, as indicated by the differences in ratings of the TU 52—70 HK controls (Tables 3 and 4). Lowest disease ratings were expressed at Decatur, MI by the K26-1, K108(3)-2, and F128(3)-1 somaclones (Table 4). Yields of the TU 52-70 HK-derived somaclones at that site ranged from 533-895 crates/acre. Plants from the bulked lines screened in Hudsonville, MI in 1993 were more compact and had less petiole twisting than TU 52-70 HK (Table 3). K108(3)-2 had more suckers than the parent cultivar but all other somaclonal lines had fewer suckers than TU 52-70 HK. Amount of pithiness was similar for all the lines. Picador rated the best for overall horticultural appearance, although it rated high for suckering (Table 3). At the Decatur site (Table 4), however, horticultural ratings varied considerably from the parent, depending upon the somaclonal line examined. Lines K128-9 and F 128(4)-9 had the worst and best overall horticultural ratings, respectively (Table 4). In 1994, the bulked R5 lines had lower, but statistically insignificant, disease levels than the parent cultivar, TU 52-70 HK (Table 5). The ratings were also statistically equivalent to the resistant control, Picador. The mean disease ratings of the somaclones was 1.3. Yields did not differ among the lines screened, except for Florida 683, which was virtually unmarketable (Table 5). Stalk shape and external horticultural appearance was the same for all the lines screened. Internal horticultural ratings were best for the SHK4 line, which were similar for Picador and Tall Utah 52-70 HK. The somaclone line SHK3 was the least acceptable for petiole interior. 49 Table 4. Reactions of R, and RS Tall Utah 52-70 HK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f.sp. apii race 2 in an unreplicated field trial near Decatur, MI, 1993. Disease Yield (55 lb Celery Line Ratingl crates/ acre) Compact2 Twisting2 Suckers2 Pithy2 F128(3)1-1 1.0 775 1.2 1.5 2.0 1.5 F128(3)1-2 1.0 720 1.5 1.5 2.0 1.2 p128(3)1-3 1.0 813 1.2 1.5 1.5 1.0 F128(3)1-4 1.1 659 1.0 1.2 1.5 1.2 F128(4)-1 1.2 780 1.2 1.0 1.5 2.0 F128(4)-2 1.1 621 1.5 1.7 2.0 1.7 F128(4)-3 1.2 879 1.2 1.5 1.2 1.5 F128(4)-4 1.1 687 1.2 1.5 1.5 2.0 F128(4)-5 1.0 791 1.2 1.2 1.5 2.0 F128(4)-6 1.0 670 1.5 1.2 1.7 1.7 F128(4)-7 1.0 599 1.2 1.5 2.0 1.7 F128(4)-8 1.1 533 1.2 1.2 2.0 1.2 F128(4)-9 1 l 807 1.0 1.0 1.2 1.5 F128(4)-10 1.1 720 1.7 1.2 1.7 1.0 F128(4)-11 1.1 786 1.2 2.0 1.5 2.0 F128(4)-12 1.1 780 1.5 2.0 1.2 1.2 K26-1-1 l 0 786 1.2 1.5 1.5 1.7 [(254-2 1.0 747 1.5 1.2 1.7 2.0 K26-1-4 1.0 780 1.7 1.2 1.7 1.7 K26-1-5 1.0 895 1.2 1.5 1.5 2.0 K26-1-8 1.1 764 1.5 1.0 1.7 1.5 K108(3)2-1 1.0 736 1.5 1.0 1.5 1.5 K108(3)2-2 1.0 802 1.5 1.2 2.0 1.5 K108(3)2-3 1.1 648 1.2 1.2 1.5 1.5 K108(3)2-4 1.0 681 1.2 1.0 1.7 1.2 [(103(3)2-5 1.0 681 1.7 1.5 1.5 1.5 K108(3)2-6 1.0 670 1.5 1.2 1.7 1.5 K128-1 1.2 714 1.2 1.2 1.5 1.0 K128-4 1.1 813 1.2 1.5 1.2 1.7 [(123-3 1.2 736 1.5 1.5 1.5 1.5 [(123-9 1.1 725 1.5 2.0 1.7 2.0 [(123-10 1.2 829 1.5 1.5 1.7 1.5 [(123-11 1.0 791 1.5 1.2 1.7 1.2 [(123-12 1.0 786 1.5 1.2 1.2 1.7 K128-14 - - 1.2 1.5 - 1.2 [(123-15 1.0 829 1.5 1.5 1.5 1.2 K128-16 1.0 736 1.5 1.2 1.7 2.0 TU 52-70 HK 1.5 725 ~ 1.5 1.5 1.2 1.5 ‘ Disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Mean value of ten plants rated for compactness, twisting, suckering, and pithiness using a rating scale of 1 to 3, where 1=most desirable and 3=least desirable. 50 Table 5. Reactions of bulk Tall Utah 52-70 I-IK-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f.sp. apii race 2 in a replicated field trial near Hudsonville, MI, 1994. Disease Yield (55 lb Celery Line Ratingl crates/acre) Stalk Shape2 External2 Internal2 SHKl (R5) 1.3 ab3 628 a3 1.3 a3 1.0 a3 1.1 ab3 SHK2 (R5) 1.3 ab 7823 1.2a 1.1a 1.2 ab SHK3 (R5) 1.4 ab 724a 1.3a 1.1a 1.4a SHK4 (Rs) 1.3 ab 732a 1.4a 1.0a 1.1 b SHKS (11,) 1.2 ab 786a 1.1a 1.0a 1.2 ab Controls: Picador 1.1a 838 a 1.5 a 1.1a 1.0 b TU52-70HK 1.8 b 591 a 1.3 a 1.0 a 1.0 b Florida 683 3.8 c 58 b NR4 NR4 NR‘ ‘ Disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Horticultural characteristics rated were stalk appearance, external markings, and internal acceptability using a rating scale of 1 to 3, where 1 = most desirable and 3 = least desirable. 2 Within columns, means followed by the same letter were not significantly different according to Tukey’s Test (P = 0.05). ‘ NR = not rated because of insufficient plants. 51 Disease ratings of the somaclones relative to the parent cultivar, Tall Utah 52-70 HK, improved following yearly selection. The R4 generation had an average relative disease rating of 0.8 and the R5 generation had an average relative disease rating of 0.73. Florida 683-derived somaclone screening In 1992, generations R1 and R2 of Florida 683-derived somaclones all had disease ratings less than the parent celery cultivar Florida 683 (Table 6). The most improvement in Fusarium yellows resistance was seen (Table 6) in somaclonal lines FLl, FL3-2, and FL3-5. FL3-2 also had the largest yield (1154 crates/acre), whereas FL3-5 had the lowest (725 crates/acre). The two most compact lines were FL3-4 and FL3-1. FL3-1 also exhibited substantial petiole twisting compared with the other lines (Table 6). FL4 and FL3-5 had the best horticultural ratings overall. In 1993, R2 and R3 generations of the selected Florida 683-derived lines were examined. Lines derived from FL4 had the worst disease ratings, although two of these lines (FLA-1 and FL4-3) had less disease than the parent cultivar Florida 683 (Table 7). Most lines exhibited better Fusarium yellows resistance compared with the Florida 683 cultivar from which they were derived. Somaclonal lines produced yields ranging from 555-912 crates/acre; Picador produced the highest yield at 1066 crates/acre (Table 7). At the Decatur site, the most horticulturally desirable lines were FL1-4, FL1-6, and FL3-4—2. Average relative disease ratings of the somaclones compared with the parent Florida 683 cultivar fell over the three years of selection. The average relative disease ratings were 0.67 for R,, 0.62 for R2, and 0.53 for R3. 52 Table 6. Reactions of Florida 683-derived celery lines to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in an unreplicated field trial near Muskegon, MI, 1992. Disease Yield (55 lb Celery Line Rating‘ crates/acre) Compact2 Twisting2 Pithy2 FLl (11,) 1.0 1060 1.1 1.0 1.0 FL3 (11,) 2.0 -3 - - - FL4 (R,) 1.1 1060 1.0 1.0 1.0 FL6 (11,) 1.5 978 1.1 1.0 1.0 FL3-l (R2) 1.4 950 1.4 1.4 1.0 FL3-2 (R2) 1.0 1154 1.0 1.0 1.0 FL3-3 (R2) 1.1 989 1.1 1.2 1.0 FL3-4 (11,) 1.1 1093 1.5 1.1 1.0 FL3-5 (R2) 1.0 725 1.0 1.0 1.0 Control: Florida 683 2.1 1000 - - - ‘ Disease ratings were based on a 1 t0 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Mean value of ten plants rated for compactness, twisting, and pithiness using a rating scale of 1 to 3, where 1=most desirable and 3=least desirable. 3 No values were available because there were only 2 plants available. 53 Table 7. Reactions of R2 and R3 Florida 683-derived celery lines and controls to Fusarium yellows in soil naturally infested with Fusarium oxysporum f. sp. apii race 2 in an unreplicated field trial near Decatur, MI, 1993. Disease Yield (55 lb Celery Line Ratingl crates/acre) Compact2 Twisting2 Suckers2 . Pithy2 FL 1-2 1.3 687 1.5 1.0 1.7 1.5 FL 1-4 1.0 846 1.2 1.0 1.0 1.0 FL 1—6 1.0 725 1.0 1.0 1.2 1.0 FL 2-1 1.0 714 1.2 1.0 2.0 1.5 FL 3-2-1 1.2 769 1.7 1.0 1.2 1.0 FL 3-2-2 1.0 692 1.5 1.2 1.7 1.0 FL 3-2-4 1.0 758 1.2 1.0 1.5 2.0 FL 3-2-5 1.0 758 1.5 1.2 1.7 1.5 FL 3-3-1 1.0 687 1.2 1.0 1.7 1.0 FL 3-3-3 1.2 780 1.5 1.0 1.2 1.0 FL 3-3-5 2.2 824 1.2 1.2 1.7 1.5 FL 34-2 1.3 912 1.5 1.0 1.0 1.0 FL 3-4-3 1.5 753 1.5 1.5 1.2 1.5 FL 4-1 1.2 753 1.5 1.2 1.7 1.7 FL 4-2 1.9 720 1.2 1.2 1.5 1.5 FL 4-3 1.0 912 1.5 1.2 1.2 1.2 FL 4-4 1.6 654 1.7 1.5 1.2 1.2 FL 4-5 3.3 555 1.7 1.2 2.0 2.0 Controls: Florida 683 2.2 747 1.7 1.5 1.5 1.7 TU52—70HK 1.5 725 1.5 1.5 1.2 1.5 Picador 1.0 1066 1.7 1.0 1.2 1.0 ' Disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2 Mean value of ten plants rated for compactness, twisting, suckering, and pithiness using a rating scale of 1 to 3, where 1=most desirable and 3=least desirable. 54 In 1994, heavy rains slowed celery maturation and damaged some plots of FL 683-derived somaclones. These factors delayed harvest so that data were not available to be included at this time. 55 DISCUSSION In 1952, the discovery of a highly Fusarium yellows-resistant selection which was later released as Tall Utah 52-70 (Elmer 1985, Hart and Endo 1975) allowed the virtual elimination of Fusarium yellows disease caused by Fusarium oxysporum f.sp. apii race 1 (Elmer 1985). The devastating arrival of Fusarium oxysporum f. sp. apii race 2 spurred researchers on to discover adequate control measures for this disease. Ideally, disease control would lead to successful eradication of the pathogen, as was seen with Tall Utah 52-70. However, in many systems host resistance adds selective pressure and encourages genetic population shifts in the pathogen that favor virulence (McDonald et al. 1989). Soilborne Fusarium oxysporum populations are unlikely to develop genetic shifts in virulence as quickly as pathogens that can spread rapidly or are polycyclic (Leonard and Fry 1986). Also, since FOA2 has competitive saprophytic capability (Opgenorth and Endo 1981), less virulent strains of the fungus may compete saprophytically between host infections as efficiently as virulent strains (Roberts 1978). Despite changes in pathogenicity, the populations are likely to remain fairly stable for long periods of time (Roberts 1978). FOA2 was also thought to have arisen from a single mutation (Toth 1989). The FOA2 population has been genetically uniform with regards to isozymes (Toth 1989), morphology (Awuah and Lorbeer 1989, Correll et al. 1985), and vegetative compatibility (Correll et al. 1985), although genetic alterations may have occurred recently in some areas (K.V. Subbarao, personal communication). Genetic similarity, low mutation rates, competitive saprophytic ability, and the physical limitations of spread may predispose FOA2 to being successfully controlled for long periods with host 56 resistance (McDonald et al. 1989) regardless of whether resistance is vertical or horizontal (Stall and Roberts 1978). The results from the field screening of the Tall Utah 52-70 HK-derived somaclones indicated that their disease resistance is heritable and stable. Improvements in average disease resistance relative to the parent cultivar, TU 52-70 HK, were seen from the R, to the R5 generations, although the differences in absolute disease levels of each line was rarely significantly different. Some of the individual lines were disease- free but the bulked lines in all years had average disease ratings between 1.0 and 2.0. In addition, some individual lines had elevated disease severity. Such lines were not chosen for seed production. In 1994 at Petoseed Company, California, celery seed (SHK) was produced from five R4 families using open pollination among members of each family. Until that point, seed from each generation had been produced by self-fertilization of individual plants. Since celery seed is normally produced with open pollination, with outcrossing rates of 50-90% (Orton and Arus 1984), it was important to discover whether open pollination affected the quality of the somaclonally derived celery. Field screens of the SHK lines indicated that disease resistance and horticultural characteristics were similar to the previous year’s ratings. Open pollination may improve the vigor of celery and provide better morphological uniformity compared with inbred populations (Orton and Arus 1984). The SHK celery lines are expected to be made available to the seed industry as disease resistant germplasm in 1995. Somaclonal variation frequently results in alterations in more than one trait (Daub 1986). Although the selection of somaclones has been based primarily on disease 57 ratings, horticulturally characteristics also have been assessed to ensure that the selected Fusarium yellows-resistant somaclonal lines would be acceptable for market. Somaclones were rated for amount of suckering, compactness, yield, petiole twisting, pithiness, and stalk shape. In previous years, brittleness, growth cracks and flavor also were examined (B.D. Cortright, personal communication, M.L. Lacy et al. unpublished). Other important traits of celery include growth rate, color, numbers of petioles, height, bolting resistance, and petiole characteristics such as length, width, thickness, and tenderness (Swiader et al. 1992). The R5 SHK somaclones had TU 52-70 HK-type characteristics; ratings of stalk shape were the same for all lines as the TU 52-70 HK parent. SHK3, however, had worse internal qualities such as hollow and more pithy petioles than the parent. Somaclones from susceptible parent cultivars are less likely to exhibit disease resistance than are somaclones derived from moderately resistant cultivars (Daub 1986, Wright and Lacy 1988). This study showed that disease resistance of somaclones derived from the highly susceptible parent cultivar Florida 683 has been successfully improved with field screening and selection. Relative disease ratings of the Florida 683-derived somaclones compared to the parent Florida 683 have improved over successive generations of screening. Florida 683, which often has a disease rating of 3.5 to 5.0 (Wright and Lacy 1988), had average disease ratings of 2.1 and 2.2 for 1992 and 1993, respectively. Low disease pressure and/or less stressful environmental conditions could cause the somaclones to appear more resistant than they are. Although selections of the most resistant Florida 683—derived somaclones improved average disease resistance, a few individual lines had fairly high levels of 58 disease. One line, FLA-5, had more disease than the highly susceptible parent Florida 683. Even some generation R3 plants from the somaclonal line furthest along in deve10pment (Figure 1) FL3, had elevated disease levels. The appearance of lines with worse disease resistance may be attributable to ”escapes" (plants that had avoided infection or symptom expression without being resistant) or a genetic change resulting in a loss of disease resistance. Loss of resistance could occur in progeny having two recessive alleles of a dominantly controlled resistance gene. This would explain the sudden appearance of progeny with very poor disease resistance following selection and self-fertilization of disease-free plants. Horticultural characteristics of the Florida 683- derived somaclones have been stable and similar to those of Florida 683. Celery somaclones regenerated from tissue culture exhibit a range of disease susceptibility, even when the parent cultivar was highly susceptible (Wright and Lacy 1988). All Florida 683-derived somaclones in this study were selected from 2 callus batches (Figure 1). Other calluses that formed produced either morphologically aberrant somaclones or regenerants with poor disease resistance so that they were not used for further selection. The plant cells used to produce callus numbers 5 and 1 (Figure 1) either had preexisting higher levels of resistance or were more susceptible to culture- induced changes that led to disease resistance (Daub 1986, Orton 1984). Although the mechanism of disease resistance is unknown, it would be likely that all plants regenerated from the same callus attained disease control through similar means. A celery somaclone derived from TU 52-70 R with Fusarium yellows resistance had 2 loci for disease resistance (Heath-Pagliuso and Rappaport 1990). Possibly, genes for resistance already existed that were silent or unexpressed (Wink 1988). Genetic 59 Original cultivar Florida 683 Callus number 5 2 1 / \ \\\\ R}, FLl FL2 FL3 FL4 / / \\ / R1 FL1-4 FL1-6 FL2-1 FL3-2 FL3-3 FL3-4 FL4-3 R2 FL1-4-1 FL1-6-1 FL2-1-1 FL3-3-1 FLA-3-1 to to to to FLl-4-8 FL1-6-5 FL2-1-4 F13-4-2 FL4-3-8 ,///' / I///// //// FL3-2-2 FL3-2-4 FL3-2-5 R3 FL3-2-2-1 FL3-2-4-1 FL3-2-5-1 FL3-3-1-1 FL3-4-2-1 to to to to to FL3-2-2-5 FL3-2-4-8 FL3-2-5-4 FL3-3-1-5 FL3-4-2-6 Figure . Lineage of Florida 683-derived celery somaclones selected for increased resistance to Fusarium yellows of celery. 1Somaclones regenerated from tissue culture called D, 5, C, and B were renamed FLl, FL2, FL3, and FL4, respectively. 6O rearrangement within cells is commonly observed in tissue culture (Murata and Orton 1984) which may allow silenced resistance genes to be expressed. Alternatively, somaclonal variation may select or induce cells to produce higher levels of secondary compounds that have antimicrobial properties. Resistant celery cultivars expressed higher levels of unidentified phenol-containing compounds (Jordan et al. 1989) as well as furanocoumarins (Cerkauskas and Chiba 1991) than susceptible cultivars. Resistance genes, therefore, may in part be responsible for the production of antifungal compounds. It has not been determined whether Fusarium yellows control in celery somaclones is provided by elevated levels of secondary metabolites such as furanocoumarins. Despite selections of plants with low disease ratings, progeny of segregating generations do not provide absolute disease control. The range of Fusarium yellows disease levels seen in field screenings of somaclones suggests that horizontal resistance was selected for rather than vertical resistance to race 1 (Stall and Roberts 1978). Resistance in celery is conferred by one major dominant gene and several independent quantitative genes (Orton et al. 1984a). The high resistance of TU 52-70 to FOAl indicated that the cultivar had vertical resistance to race 1 (Stall and Roberts 1978). Although vertical resistance may be preferable for FOA2 control, current information indicates that it may be difficult to achieve. Horizontal resistance usually operates by lowering the rate of infection by the pathogen and reducing onset of disease (Stall and Roberts 1978). In contrast to vertical resistance, horizontal resistance may not significantly lower population levels of the pathogen in the field (Stall and Robert 1978). Eradication of FOA2 with celery somaclones or other cultivars with moderate resistance may not be possible unless vertical resistance is developed. Because expression of 61 horizontal resistance is affected by environmental factors (Stall and Roberts 1978) and horizontal resistance does not completely prevent disease, using host resistance for the sole disease management may not be possible. It is recommended that other control measures are used in combination with planting of celery lines with good Fusarium yellows resistance. LITERATURE CITED Awuah, R.T., J .W. Lorbeer, and A. Ellerbrock. 1986. Occurrence of Fusarium yellows of celery caused by Fusarium oxysporum f. sp. apii race 2 in New York and its control. Plant Dis. 70:1154-1158. Awuah, R.T. and J.W. Lorbeer. 1989. Role of light, temperature, and method of propagation in cultural variability of Fusarium oxysporum f. sp. apii race 2. Mycol. 81:278-283. Bank-Izen, M.S., T.J. Orton, and L. Rappaport. 1981. Fusarium-induced toxin produced in celery and effects on celery cells in suspension culture. HortSci. 16:457. (Abstr.) Becker, J.O., C.A. Hepfer, G.Y. Yuen, S.D Van Gundy, M.N. Schroth, J.G. Hancock, A.R. Weinhold, and T. Bowman. 1990. Effect of rhizobacteria and metham-sodium on growth and root microflora of celery cultivars. Phytopathol. 802206-211. Brettell, R.I.S. and D.S. Ingram. 1979. Tissue culture in the production of novel disease-resistant crop plants. Biol. Rev. 54:329-345. Browers, M.A. and T.J. Orton. 1982a. A factorial study of chromosomal variability in callus cultures of celery (Apium graveolens). Plant Sci. Lett. 26:65-73. Browers, M.A and T.J. Orton. 1982b. Transmission of gross chromosomal variability from suspension cultures into regenerated celery plants. J. Heredity 73:159-162. Cerkauskas, R.F. and M. Chiba. 1991. Soil densities of Fusarium oxysporum f.sp. apii race 2 in Ontario, and the association between celery cultivar resistance and photocarcinogenic furocoumarins. Can. J. Plant Pathol. 13:305-314. Correll, J .C., J.E. Puhalla, and R.W. Schneider. 1985. Distinct vegetative compatibility groups (populations) of Fusarium oxysporum colonizing celery roots from California. Phytopathol. 75:1347. (Abstr.) 62 63 Daub, ME. 1986. Tissue culture and the selection of resistance to pathogens. Ann. Rev. Phytopath. 24: 159-186. Elmer, W.H. 1985. The ecology and control of Fusarium yellows of celery in Michigan. PhD. Dissertation. Michigan State University, East Lansing, MI. p. 146. Elmer, W.H. and M.L. Lacy. 1984. Fusarium yellows (F. oxysporum f.sp. apii) of celery in Michigan. Plant Dis. 68:537. (Abstr.) Elmer, W.H., M.L. Lacy, and S. Honma. 1986. Evaluations of celery germ plasm for resistance to Fusarium oxysporum f.sp. apii race 2 in Michigan. Plant Dis. 70:416-419. Evans, D.A., W.R. Sharp, and HP. Medina-Filho. 1984. Somaclonal and gametoclonal variation. Am. J. Bot. 71:759-774. Hart, LP. and R.M. Endo. 1975. Fusarium yellows of celery in California. Pro. Am. Phytopath. Soc. 2:114. Hartman, C.L., T.J. McCoy, and TR. Knous. 1984. Selection of alfalfa (Medicago sativa) cell Lines and regeneration of plants resistant to the toxin(s) produced by Fusarium oxysporum f.sp. medicaginis. Plant Sci. Lett. 34:183-194. Heath-Pagliuso, S. and L. Rappaport. 1990. Somaclonal variant UC-T3: the expression of Fusarium wilt resistance in progeny arrays of celery, Apium graveolens L. Theor. Appl. Genet. 80:390-394. Heath-Pagliuso, S. , J. Pulhnan, and L. Rappaport. 1988. Somaclonal variation in celery: screening for resistance to Fusarium oxysporum f. sp. apii. Theor. Appl. Genet. 75:446-451. Heath-Pagliuso, S., J. Pullman, and L. Rappaport. 1989. ’UC-T3 somaclone’: celery germplasm resistant to Fusarium oxysporum f.sp. apii., race 2. HortSci. 24:711-712. Honma, S. and M.L. Lacy. 1980. Hybridization between pascal celery and parsley. Euphyt. 29:801-805. Honma, S., M.L. Lacy, and H.H. Murakish. 1986. ’Advantage’, ’Pilgrim’, and ’Companion’ celery. HortSci. 21:1073-1074. Ireland, K.F., W.H. Elmer, and M.L. Lacy. 1987. Field evaluation of celery germplasm for resistance to Fusarium oxysporum f.sp. apii race 2. Phytopathol. 7721712. (Abstr.) 64 Jordan, C.M., L.S. Jordan, and R.M. Endo. 1989. Association of phenol-containing structures with Apium graveolens resistance to Fusarium oxysporum f. sp apii race 2. Can. J. Bot. 67:3153-3163. Kallak, HI. and M.A. Vapper. 1985. Plant tissue culture as a model system for mutagenicity testing of chemicals. Mutat. Res. 147251-57. Larkin, P]. and W.R. Scowcroft. 1981. Somaclonal variation - a novel source of variability from cell cultures for plant improvement. Theor. Appl. Genet. 60:197-214. Leonard, KI. and WE. Fry. 1986. Plant disease epidemiology: pepulation dynamics and management. Macmillan Publishing Co. New York, NY. p.372. McDonald, B.A., J.M. McDermott, S.B. Goodwin, and R.W. Allard. 1989. The population biology of host-pathogen interactions. Ann. Rev. Phytopath. 27:77-94. Murata, M. and T.J. Orton. 1983. Chromosome structural changes in cultured celery cells. In Vitro Cell. Dev. Biol. 19:83-89. Murata, M. and T.J. Orton. 1984. Chromosome fusions in cultured cells of celery. Can. J. Genet. Cytol. 26:395-400. Opgenorth, D.C. and R.M. Endo. 1979. Sources of resistance to Fusarium yellows of celery in California. Plant Dis. Rep. 63:165-169. Opgenorth, D.C. and R.M. Endo. 1981. Competitive saprophytic ability (CSA) of Fusarium oxysporum f. sp. apii. Phytopathol. 71:246-247. (Abstr.) Opgenorth, D.C. and R.M. Endo. 1985. Additional sources of resistance to race 2 of Fusarium oxysporum f.sp. apii. Plant Dis. 69:882-884. Orton, T.J. 1982. Effect of Fusarium oxysporum f. sp. apii on seed germination in celery (Apium graveolens). Can. J. Bot. 60:34-39. Orton, T.J. 1983. Spontaneous electrophoretic and chromosomal variability in callus cultures and regenerated plants of celery. Theor. Appl. Genet. 67:17-24. Orton, T.J. 1984. Genetic variation in somatic tissues: method or madness? Adv. Plant Pathol. 2:153-189. Orton, T.J. and P. Arus. 1984. Outcrossing in celery (Apium graveolens). Euphyt. 33:471-480. Orton, T.J., M.E. Durgan, and SD. Hulbert. 1984a. Studies on the inheritance of resistance to Fusarium oxysporum f.sp. apii in celery. Plant Dis. 68:574-578. 65 Orton, T.J., S.H. Hulbert, M.E. Durgan, and CF. Quiros. 1984b. UCl, Fusarium yellows-resistant celery breeding line. HortSci. 19:594. Otto, H.W., A.O. Paulus, M.J. Snyder, R.M. Endo, L.P. Hart, and J. Nelson. 1976. A crown rot of celery. Cal. Agr. 30210-11. Pullman, G. and L. Rappaport. 1983. Tissue culture-induced variation in celery for Fusarium yellows. Phytopathol. 73:818. (Abstr.) Pullman, G., L. Rappaport, and S. Heath-Pagliuso. 1984. Somaclonal variation in celery: towards selection for resistance to Fusarium wilt. HortSci. 19:589. (Abstr.) Roberts, DA. 1978. Fundamentals of plant-Est control. WH Freeman and Co., San Fransisco, CA. p. 242. Schneider, R.W. 1984. Effects of nonpathogenic strains of Fusarium oxysporum on celery root infection by Fusarium oxysporum f. sp. apii and a novel use of the Lineweaver-Burk double reciprocal plot technique. Phytopathol. 74:646-653. Schneider, R.W. 1985. Suppression of Fusarium yellows of celery with potassium, chloride, and nitrate. Phytopathol. 75:40-48. Stall, RE. and DA. Roberts. 1978. Plant pathology and the control of plant diseases. pp. 89-121. IN Fundamentals of plant-pest control. D.A. Roberts ed. WH Freeman and Co., San Fransisco, CA. p. 242. Swiader, J .M., G.W. Ware, and J.P. McCollum. 1992. Producing vegetable crops. Fourth ed. Interstate Publishers Inc., Danville, IL. p.626. Toth, K.F. 1989. Biology and control of Fusarium oxysporum f. sp. apii race 2. PhD. Dissertation. Michigan State University, East Lansing, MI. p.174. Toth, K.F and M.L. Lacy. 1991. Increasing resistance in celery to Fusarium oxysporum f. sp. apii race 2 with somaclonal variation. Plant Dis. 75: 1034-1037. Wink, M. 1988. Plant breeding: importance of plant secondary metabolites for protection against pathogens and herbivores. Theor. Appl. Genet. 75:225-233. Wright, J .C and M.L. Lacy. 1988. Increase of disease resistance in celery cultivars by regeneration of whole plants from cell suspension cultures. Plant Dis. 72:256-259. CHAPTER 3 - EVALUATION OF SOIL AMENDMENTS AND COVER CROPS FOR THEIR EFFECT ON DISEASE EXPRESSION INTRODUCTION Soilborne diseases are persistent and difficult to control in the field. Management of soilbome diseases usually involves crop rotation, host resistance, and may include soil fumigation (Agrios 1988). Fusarium yellows of celery, caused by Fusarium oxysporum f. sp. apii race 2 (FOA2), has caused large losses in celery yield and quality (Becker et al. 1990) since its appearance in the 1970’s (Otto et a1. 1976). In Michigan this disease is the limiting factor in celery production because reliable controls have not been available (T 0th and Lacy 1991). Soil fungicidal drenches have not provided economical control of Fusarium yellows of celery (Awuah et al. 1986, Otto et al. 1976, M.L. Lacy personal communication). Methyl bromide, a soil fumigant, can eradicate FOA2 from greenhouse soils and localized field sites (Awuah and Lorbeer 1991, Awuah et al. 1986) but it is economically prohibitive for entire fields (Otto et al. 1976). As well, methyl bromide may soon lose its registration for use as a soil fumigant. In searching for methods to manage Fusarium yellows of celery, modifications to the chemical and biological composition of soil were attempted. Alkaline soils were less conducive to Fusarium yellows of celery than soils with low pH (Opgenorth and 66 67 Endo 1983). Ammonium-amended soils were more conducive to disease than non- amended soils, but adding nitrate reduced Fusarium yellows of celery (Schneider 1985). Adding chitin to soil has decreased some soilbome diseases (Mitchell and Alexander 1961a, 1961b) but it did not help control Fusarium yellows of celery in a field trial (M.L. Lacy, personal communication). Incorporation of other microbes into FOA2- infested soils has been examined for Fusarium yellows control. Antagonistic bacteria helped to reduce disease in laboratory studies (Opgenorth and Endo 1983) but rhizobacteria did not reduce disease severity ratings at harvest when tested in the field (Becker et al. 1990). Nonpathogenic Fusarium oxysporum isolates from a Fusarium yellows-suppressive soil lowered celery disease severity following their incorporation into infested soil (Schneider 1984). However, none of these putative biocontrol agents are currently commercially feasible. Instead, these findings provided evidence that indigenous microorganisms could help to reduce disease levels if their populations could be selectively increased (Schneider 1984, Opgenorth and Endo 1983). Cover crops and organic crop residues may provide a way to increase activity of these types of microbes. Organic soil amendments have affected disease directly by affecting pathogen growth and dormancy and indirectly by stimulating the growth of other microorganisms that may compete with or antagonize the pathogen (Schroth and Hildebrand 1964). Since chlamydospores were determined to be the primary inoculum for most soilbome Fusarium diseases (Price 1984), research was undertaken to determine the factors regulating germination and formation of chlamydospores. It was expected that once such factors were elucidated, soil amendments could be used to control disease by causing chlamydospores to break dormancy in an environment unfavorable for chlamydospore 68 formation (Papavizas and Lumsden 1980), rendering them more susceptible to lysis (Patrick and Toussoun 1970, Sewell 1970). Chlamydospore germination is affected by plant residues, root exudates, and soil microflora, as well as abiotic factors such as ph, temperature, CO2 concentration and soil moisture (Griffin 1981, Schroth and Hendrix Jr. 1962, Sequeria 1962, Sewell 1970, Smith and Snyder 1972). Fusarium chlamydospores germinated from such nutrient stimuli as glucose and asparagine, fresh plant tissue, and host and nonhost root exudates (Garrett 1970, Schroth and Hendrix Jr. 1962, Toussoun et al. 1969), although plants within the same family as the host were more likely to stimulate spore germination (Reyes and Mitchell 1962). Lewis and Papavizas (1977) found that plant residues with a high C:N ratio (mature plants) inhibited chlamydospore germination of Fusarium solani f.sp. phaseoli but this did not hold for all Fusarium species (Huber and Watson 1970, Oritsejafor and Adeniji 1990). The bacterial flora in soil can also affect germination. Although nutrient competition with soil bacteria has prevented germination, many of the bacteria present were fluorescent Pseudomonas spp. (Schroth and Hancock 1982). These bacteria produced siderophores that complexed iron and inhibited growth of root pathogens (Schroth and Hancock 1982). Direct correlations have been found between siderophore production and inhibition of chlamydospore germination of Fusarium oxysporum f.sp. cucumerinum (Sneh et al. 1984). Chlamydospore germination may have several results. The germ tube may infect a host (Nelson et al. 1981), produce new, daughter chlamydospores (Smith and Snyder 1972, Garrett 1970), or lyse (Toussoun et al. 1969). Exogenous stimuli have determined the fate of the germ tube. Glucose and asparagine induced Fusarium germ tubes to 69 produce new spores, but forest soil extracts caused germ tubes to lyse (Toussoun et al. 1969). In the presence of many crop residues and plant rhizospheres, Fusarium germination was often non-specific, although the effects of this germination differed (Schippers and van Eck 1981, Toussoun et al. 1969). Plant rhizospheres caused rapid germ tube lysis of Fusarium oxysporum f. sp. elaeidis, but leguminous cover crops encouraged the production of daughter chlamydospores (Oritsejafor and Adeniji 1990). In the presence of a host, germinating chlamydospores caused infection which lead to disease development (Linderman 1970). Certain soil amendments encouraged selective lysis of pathogen germ tubes (Schroth and Hendrix Jr. 1962), resulting in lower levels of disease. High levels of carbon and nitrogen promoted high levels of germination but also created conditions (probably high bacterial levels) that lysed the germlings (Garrett 1970, Patrick and Toussoun 1970). Lysis of germ tubes caused by other soil microbes was called heterolysis; autolysis occurred when the carbon nutrients were exhausted (Garrett 1970). Amendments with low C:N ratio caused autolysis through carbon starvation (Garrett 1970) but those with a high C:N ratio allowed chlamydospores to form even after the nitrogen source was exhausted. Fusarium chlamydospores often form in hyphae that have colonized plant tissue, although chlamydospores also form in germ tubes following a short-lived germination period (Schippers and van Eck 1981). These replacement chlamydospores allow Fusarium to survive long periods of harsh conditions in the soil. Abiotic factors such as temperature, carbon dioxide concentration, and pH have played roles in inducing chlamydospore formation in soil (Schippers and van Eck 1981). Soil amendments with 70 a low C:N ratio favored chlamydospore formation in Fusarium oxysporum (Price 1984) possibly because available organic carbon was quickly depleted (Schippers and van Eck 1981). Evidence existed that soil bacteria triggered the formation of chlamydospores in Fusarium (Park 1954). The seasonal fluctuations of chlamydospore soil populations was also a response to a changing microbial population (Price 1984). The metabolites released by soil microflora play a large role in stimulating chlamydospore formation of Fusarium and so does energy depletion of the inoculum (Shippers and van Eck 1981). The Optimum conditions for the in vitro formation of Fusarium oxysporum f .sp. apii race 2 chlamydospores were: pH 7.1, 27°C for 5-15 days followed by 24°C at day 15, matric potential of —25 centibars, and osmotic potential of 0 bars (Opgenorth and Endo 1985). Pathogen population has not necessarily been well-correlated with disease expression. Very low inoculum densities have usually produced low levels of disease but lowering field inoculum density from very high to moderately high did not significantly reduce disease pressure (Johnson 1994). Fields with the highest inoculum levels did not necessarily produce the worst disease (Nash and Snyder 1962), although there was a very strong positive relationship between inoculum density and disease severity with Fusarium yellows of celery in Michigan (Elmer and Lacy 1987b). Also, soil treatments have reduced disease severity without lowering population levels of the pathogen (Jarvis and Thorpe 1981, Burpee 1990), suggesting the presence of fungistatic compounds (Ramirez- Villapudua and Munnecke 1988, Zacharia 1978) or changes in the susceptibility of host roots (Sun and Huang 1985). For Fusarium yellows of celery, Elmer and Lacy (1987b) found that increased levels of Fusarium oxysporum f.sp. apii race 2 in soil related to increased disease severity and decreased celery yield. This finding suggests that lowering 71 propagule levels of FOA2 in soil would control disease. However, the population level must be reduced to an inoculum density below 42 propagules per gram of soil to prevent vascular discoloration in 6 weeks (Elmer and Lacy 1987b). Fusarium root rot of bean has been the model system used to study the effects of soil organic amendments for the control of Fusarium. Mature (high C:N ratio) plant residues of rye (Secale cereale L.), oat (Avena sativa L.), soybean (Glycine max L.), barley (Hordeum vulgare L.), buckwheat (Fagopyrum esculentum (Mill.) Moench), and timothy (Phleum pratense L.) were all found to lessen disease severity of root rot of bean (Lewis and Papavizas 1975, 1977). Mature rye was most efficient for disease control, possibly because of the release of fungitoxic compounds during decomposition (Lewis and Papavizas 1975, Menzies and Gilbert 1967). Mature corn was found to only slightly reduce disease severity (Lewis and Papavizas 1975, 1977). Immature corn (Zea mays L.) and sorghum (Sorghum bicolor (L.) Moench) (low C:N ratio) were consistently able to reduce disease (Lewis and Papavizas 1975, 1977) but immature rye, oat, and buckwheat residues successfully reduced Fusarium root rot of bean in only one of two studies (Lewis and Papavizas 1975). Since maturity of plant tissue alters its C:N ratio, it was suggested that this factor may play a role in providing disease control (Schroth and Hildebrand 1964). Although mature (high C:N) residues tend to control soilbome disease better than immature (low C:N) amendments (Papavizas et al. 1968), no definite correlation between C:N ratio and disease rating has been found (Huber and Watson 1970). Amendments with high C:N ratios are thought to encourage immobilization of nitrogen (Schroth and Hildebrand 1964) which may provide disease control by discouraging pathogen germination and growth (Cook and Schroth 1965). Bean root rot 72 has been controlled by amending infested soil with cellulose (Adams et al. 1968a) but this control has been negated by the addition of exogenous nitrogen (Papavizas et a1. 1968). The action of nitrogen on the pathogen is not known. It has been suggested that nitrogen applications stimulated chlamydospore germination (Cook and Schroth 1965), enabling root infection to occur. However, chlamydospore germination in the absence of the host has also been shown to encourage lysis of germ tubes (Schroth and Hendrix Jr. 1962, Patrick and Toussoun 1970) and inorganic nitrogen has not affected chlamydospore germination in soil (Adams et al. 1968a). If nitrogen does encourage both propagule germination and disease increases, it seems likely that the nutrients allow the fungus to produce new chlamydospores that incite disease (Schroth and Hendrix Jr. 1962). Although nitrogen content could not completely explain the effectiveness of some organic soil amendments (Lewis and Papavizas 1977), high C:N organic amendments often provided some Fusarium root rot control in bean. These residues increased microbial activity in the rhizosphere of beans more readily than immature plant residues (Papavizas et al. 1968). Nonpathogenic microorganisms have reduced disease severity of many soilbome pathogens (Huber and Watson 1970, Lewis and Papavizas 1975, Mitchell and Alexander 1961b, Schroth and Hancock 1982). Soilborne bacteria induce chlamydospore formation (Ford et al. 1970) and hamper chlamydospore germination (Cook and Schroth 1965) of Fusarium solani f. sp phaseoli. Disease-controlling soil amendments may have maintained high enough populations of these bacteria to keep chlamydospores from breaking dormancy, rendering them unable to infect the host. Organic amendments may also help to control Fusarium root rot of bean by providing nutrient competition between the pathogen and the nonpathogens (Adams et al. 1968b) 73 or by increasing the populations of soil microbes antagonistic to Fusarium solani f. sp. phaseoli (Papavizas et al. 1968). Studies have been done on the effect of amendments on the control of soilbome diseases caused by formae speciales of F usarium oxysporum, including Fusarium yellows of celery. Some Fusarium wilts have been controlled with composts. Composted sewage sludge reduced Fusarium wilts of melon and cucumber by increasing the populations of microbial antagonists (Hoitink and Fahy 1986, Lumsden et al. 1983). Several Fusarium diseases have been moderately controlled with larch bark composts (Hoitink and Fahy 1986). Waste products have also been a good source of soil amendments because they are inexpensive and readily available. Infested soil amended with chicken manure (25 tons/acre-foot) has controlled Fusarium yellows of celery in greenhouse trials (R.W. Schneider unpublished). Sun and Huang (1985) developed a commercially available soil amendment for the control of Fusarium wilts. This "S-H" mixture contains wilt suppressive waste products (sugar cane bagasse, rice husks, oyster shell, and mineral ash) as well as fertilizers (Sun and Huang 1985). Disease control is provided by increased soil pH, increased antagonistic microbial activity, and calcium supplementation (Sun and Huang 1985). Oilseed meals have been good waste materials for the control of certain soilbome diseases. Severity of Fusarium yellows of celery has been lowered by amending soil with 10 tons/acre-foot of cottonseed meal (R.W. Schneider unpublished). Cottonseed and linseed meals reduced Fusarium populations in soils within 4 weeks at rates as low as 1% (w/w), although the effect was less dramatic in organic soils (Zakaria and Lockwood 1980). Oilseed meal amendments do not stimulate chalmydospores to germinate. Instead they kill dormant Fusarium spores with toxic compounds released 74 during decomposition (Zacharia 1978). These volatile toxins did not have a deleterious effect on Fusarium populations unless the oilseed-amended soil was kept in closed containers (Zacharia 1978, Zacharia and Lockwood 1980). For field use, oilseed- amended soil must be tarped to provide disease control. Crop residues have not been successful so far at controlling Fusarium yellows of celery but some may have promise. Celery residues consistently increased disease severity and celery extracts were found to promote very little spore germination (Elmer and Lacy 1987a). Onion residues, on the other hand, inhibited disease development well and onion extracts caused 62% of Fusarium oxysporum f. sp. apii chlamydospores to germinate (Elmer and Lacy 1987a). Germination of dormant propagules of FOA in the absence of the host may result in lysis of the germ tube and population reduction, although no population differences existed among soils amended with different residues after 3 weeks (Ehner and Lacy 1987a). Onion and mint amendments improved the dry weight of celery (Elmer 1985) but did not reduce disease ratings in the field (T 0th 1989). Amending soil with rye, sudax, cabbage, and brown mustard incited more disease than in unamended soil (Elmer and Lacy 1987a, Toth 1989). Cover cropping with cats and mustard each provided some disease control in the greenhouse (Toth 1989). Rye and barley cover crops did not reduce disease severity (Toth 1989). Crucifers have good potential as soil amendments for disease control of Fusarium yellows of celery. These plants belong to the family Brassicaceae and produce a battery of microbial toxins upon decomposition. Glucosinates, which range in levels from 86 to 1240 umol/100 g in crucifers, liberate isothiocyanates and thiocyanates during tissue breakdown (Carlson et a1. 1987). Isothiocyanates were not detected in air above soils 75 amended with crucifers, possibly because these compounds adsorbed to soil particles (Lewis and Papavizas 1970). Decomposing crucifers in natural soil also release methanethiol, dirnethyl sulfide, dimethyl disulfide, methanol, ethanol, acetone, and acetaldehyde (Lewis and Papavizas 1970). Other cruciferous volatiles that may be responsible for disease control (Lewis and Papvizas 1971, Ramirez-Villapudua and Munnecke 1987) were only released if the residues were heated during decomposition (Lewis and Papavizas 1970). Upon heating, isothiocyanates broke down into hydrogen sulfide and carbon disulfide (Lewis and Papavizas 1970). Cabbage has been used successfully for the control of cabbage yellows, caused by Fusarium oxysporum f. sp. conglutinans (Ramirez-Villapudua and Munnecke 1986, 1987). During decomposition, cabbage releases volatile compounds that are fungistatic to pure cultures of F. oxysporum f.sp. conglutinans but are fungitoxic in naturally infested soil (Ramirez-Villapudua and Munnecke 1986). After 15 days of exposure to volatiles produced from decomposing cabbage, soil populations of F. oxysporum f. sp. conglutinans were reduced to undetectable levels (Ramirez-Villapudua and Munnecke 1988). Following residue incorporation, fields had to be covered with a polyethylene tarp to prevent the fungitoxic volatiles from escaping in order to provide disease control (Ramirez-Villapudua and Munnecke 1986). Polyethylene tarps also raise soil temperature by solarization (solar heating) (Katan 1981, Ramirez-Villapudua and Munnecke 1986) which, when used with dried cabbage residues, nearly eliminated cabbage yellows (Ramirez-Villapudua and Munneck 1987). Dried residues, elevated residue incorporation rates, and longer solarization periods all have provided better disease control (Ramirez-Villapudua and Munneck 1986, 1988). Also the color, thickness, and optical properties of the 76 polyethylene tarp affected the maximum soil temperatures obtained (Ham et a1. 1993, Katan 1985). Control of cabbage yellows has also been achieved with dried residues of cabbage, cauliflower, broccoli, collard, mustard, and kale amended to soil and tarped (Ramirez-Villapudua and Munnecke 1988). All of these crucifers gave even better disease control if the amended and tarped soil was also solarized (Ramirez-Villapudua and Munnecke 1988). Solarization alone has also been used to control soilbome diseases (Katan 1985). Soil disinfestation of soilbome pathogens has been achieved by tarping moistened soil with transparent polyethylene for several weeks (Katan 1981, 1985). These mulches (tarps) allow solar radiation to heat the soil to temperatures of 40-60°C if the site has intense solar irradiation (Katan et al. 1976, Hardy and Sivasithamparam 1985). Although a few Fusarium diseases have not been satisfactorily suppressed by solar heating, Fusarium disease of cotton, melon, tomato, onion, and strawberry have all been controlled with solarization (Katan 1985). Solarization heats soils at elevated but sublethal temperature for several weeks (Katan 1985). This slow heating has weakened pathogen propagules and made them more susceptible to degradation by external stresses (Katan 1985). Mulches prevented gases from escaping that adversely affected pathogen growth, competitive ability, or virulence that accounted for disease control (Katan 1985). For instance, the 35-fold increase in carbon dioxide concentration observed in tarped soils compared with open soils (Katan 1985) may have been responsible for disease suppression (Patrick and Toussoun 1970). Lengthy soil heating may add selective pressure in favor of heat-resistant saprophytes that can efficiently compete with the pathogen (Katan et al. 1976). 77 The effectiveness of solarization is affected by the optical properties of the tarp used (Ham et al. 1993), intensity of solar radiation (Katan 1981), and the length of treatment (Katan et al. 1976). Preferably, a thin (0.03 mm), transparent, polyethylene covering (Katan et a1. 1976) should be used for at least four weeks in locations where daytime soil temperatures of 40-60°C can be achieved (Katan 1981, 1985). However, occasionally soilbome pathogens have been controlled in cooler climates using solarization alone or in combination with other control measures (Katan 1985). The objective of this research was to determine which crops would most likely reduce the incidence of Fusarium yellows of celery when used in cover cropping, crop rotation, or residue incorporation in the field. This research assessed the ability of rye, cabbage, mustard, rape, and celery cover crops and rye, cabbage, mustard, rape, onion, and celery soil amendments to lower the incidence of Fusarium yellows of celery in the greenhouse. It also examined the effectiveness of soil heating in combination with mustard, cabbage or onion soil amendments in reducing Fusarium yellows of celery. 78 MATERIALS AND METHODS Inoculum Production A modification of Ehner’s (1985) soil infestation technique was used to produce Fusarium oxysporum f. sp. apii race 2 inoculum. Wheat straw was ground with a Wiley Mill and passed through a 3 mm mesh screen. Fifty grams of chopped straw were mixed with 100 ml of 0.025 M L—asparagine in an Erlenmeyer flask and sterilized. The sterile medium was inoculated with three 7 mm diameter PDA plugs that were colonized with Fusarium oxysporum f. sp. apii race 2. PDA plates were inoculated with colonized sterile muck soil stored at 2°C. Inoculated flasks were incubated at 23 j: 2°C under fluorescent lights with a 12 hour photoperiod. Flasks were shaken periodically to distribute the mycelium and encourage better colonization of the straw. In 3 weeks, the colonized straw was dried in the fume hood and stored at room temperature. Selection of Best Soil Infestation Method Several variables were examined for their effect on disease severity of Fusarium yellows of celery in order to increase the disease levels observed in the greenhouse studies. Two different Fusarium oxysporum f. sp. apii race 2 Michigan isolates were used. The Whitehall isolate was from Hekkema Farms, Whitehall, MI and the Hudsonville isolate was from Superior Brand Produce, Hudsonville, MI. Both isolates were originally isolated from infected celery in 1982 and were single-spored in 1983. Colonized wheat straw of each of these isolates was infested into soil at a low rate (3.08 g/kg) and at a high rate (6.18 g/kg). Different soils were examined as well. Naturally 79 FOA2-infested muck soil from VerHage Farms, Hudsonville, MI; FOA2-uninfested muck soil from the MSU Muck Farm, Michigan State University, Bath, MI; and Baccto" professional potting mix (Michigan Peat Company) were each used alone or with additions of FOA2-colonized wheat straw. To determine if pasteurized soil improved soil colonization by FOA2, F OA2-uninfested muck soil from the Muck Farm, Michigan State University, MI was steamed for 1 hour and infested with either dried or undried inoculum. A total of 20 treatments were assessed for their ability to produce high disease levels. For each treatment, 4 plastic bags (replicates) were each filled with 0.45 kg of soil and the appropriate amount of inoculum (if any was used). Each bag was shaken well to mix inoculum with the soil and the infested soil was evenly distributed between two 10 cm plastic pots. Two one-month—old Florida 683 celery seedlings were transplanted into each pot. In one pot, seedlings were planted at a shallow depth, with the crown at the soil line. In the second pot, seedlings were planted with the crowns 2-3 cm below the soil line. Pots were arranged on greenhouse benches in a completely randomized design. Plants were watered daily, exposed to natural light supplemented with 400 Watt high pressure sodium vapor lamps with a 16 hour photoperiod. Six weeks following transplanting, celery plants were weighed and rated for disease by examining vascular discoloration of the crowns, where 1 = no vascular discoloration, 2 = trace of vascular discoloration, 3 = < 50% crown area discolored, 4 = 2 50% crown area discolored, and 5 = dead or nearly dead plant (Toth 1989). 80 Disease severity values were each averaged per pot and these data were analyzed with a 2-way ANOVA procedure using PC-SAS software (SAS Institute Inc., Carry, NC), where one factor was depth of planting and the other factor was soil composition (isolate, inoculum level, and soil used). Tukey’s test (P=0.05) was used to detect mean separation among all treatments. Correlations between celery disease and fresh weights for the whole experiment and within each soil type were performed using PC-SAS (SAS Institute Inc., Carry, NC). For each of these correlations, the steamed treatments were not included. The results from this test determined which soil infestation method produced the highest level of disease which was used for the soil amendment and cover crop experiments. Soil Amendments Fresh tissue was collected from 4-week-old forage rye (Secale cereale L.), mustard (Brassica juncea (L.) Czernj. & Coss), winter rape (Brassica napus L.), and hybrid Spartan Banner onion (Allium cepa L.) and from 2-4 month-old Florida 683 celery (Apium graveolens L. var. dulce (Mill.) Pers.) and Tuffy hybrid cabbage (Brassica oleracea L.). Tissue was chopped with scissors into pieces approximately 1 cm2. Wheat straw colonized by the Hudsonville isolate was added to naturally infested muck soil from VerHage Farms, Hudsonville, MI at a rate of 3.08 g inoculum/kg soil. The soil and inoculum were mixed in a concrete mixer for 20 minutes. Plant tissue and artificially infested soil were weighed and mixed separately for each pot (replicate) at a rate of 2% (w/w). Naturally infested soil and soil steamed for 1 hour were used as infested and 81 noninfested controls, respectively. Fifteen replicates were used for each treatment in order to detect a true effect, based on preliminary results (Gill 1993, Gilligan 1986). Five greenhouse locations were selected and 3 pots per treatment were placed at each location. Pots were arranged randomly at each location. Soil was lightly watered daily for two weeks to encourage amendment decomposition. Two weeks after amending soil, 2 one-month-old Florida 683 celery seedlings were transplanted into each pot with the crowns 2-3 cm below the soil line. Plants were watered daily and fertilized every 2 or 3 weeks with 20-40 ml, 65 ppm (4.9 g/I) Peters‘1° 20-20-20 (N-P-K) soluble fertilizer. Natural lighting was supplemented with 400 Watt high pressure sodium vapor lamps with a 16 hour photoperiod. Daytime greenhouse temperatures ranged from 20- 30°C. The experiment was performed twice. Eight weeks following celery transplanting, plants were cut diagonally through the crown and rated for Fusarium yellows of celery based on a scale, where 1 = no vascular discoloration, 2 = trace of vascular discoloration, 3 = < 50% crown area discolored, 4 = 2 50% crown area discolored, and 5 = dead or nearly dead plant (T 0th 1989). Soil Heating and Mulching To determine if tarping and heating helped disease control, pots containing soil amended with cabbage, mustard, and onion as well as unamended and noninfested controls were produced as described above. Fifteen pots per treatment were watered lightly and wrapped in plastic bags and placed 15 cm below 400 Watt high pressure sodium vapor lamps with a 16 hour photoperiod, for 2 weeks. Daytime temperatures of 82 soil 8-10 cm below the soil line ranged from 38—48°C. Nighttime soil temperatures fell to 25 j; 2°C. After the 2 week heating period, pots were removed from the bags and 2 one-month-old Florida 683 celery seedlings were transplanted per pot. Pots were arranged on greenhouse benches as a randomized block design under the same conditions described for the amendment work without tarping. The experiment was performed twice. Celery plants were rated for Fusarium yellows as described above 8 weeks after transplanting. Cover Crops Naturally infested VerHage muck soil was supplemented with 3.08 g wheat straw colonized by the Hudsonville isolate per kg soil; 0.23 kg artificially infested soil was placed in each 10 cm plastic pot. Rye, mustard, and rape were directly seeded (5-7 seeds per pot) into infested soil. For cabbage and celery, 2 one-month-old seedlings were transplanted into each pot. Onion seed germination was so low that it could not be used. For each cover crop, 15 pots (replicates) were used, based on preliminary results (Gill 1993, Gilligan 1986). At each of 5 greenhouse locations, 3 pots per treatment were arranged in a randomized design. Natural light was supplemented with 400 Watt high pressure sodium vapor lamps with a 16 hour photoperiod and daytime greenhouse temperatures ranged from 20-30°C. Cover crop seedlings were watered daily. Following a 2 week growing period, cover crops (root tissue and foliage) from each pot were cut into small pieces, mixed with the soil and replaced. Soil was watered daily for two weeks to encourage cover crop decomposition. Two weeks after incorporating the cover crop into soil, 2 one-month-old Florida 683 celery seedlings were transplanted into 83 each pot with the crowns 2-3 cm below the soil line. Celery plants were watered daily and fertilized every 2 or 3 weeks with 20—40 ml, 65 ppm (4.9 g/l) Peters‘1° 20-20-20 (N- P-K) soluble fertilizer. Greenhouse conditions were as described above. The experiment was performed twice. Celery plants were rated for Fusarium yellows as described above 8 weeks following transplanting. Statistical Analysis For the amendment experiment (with and without tarping), average disease ratings per pot were analyzed with a 2-way ANOVA using SigmaStat (Jandel Scientific Software, San Rafael, CA), testing for treatment and block main effects as well as treatment and block interactions. Contrasts were chosen to compare: 1) the unamended control with each amended treatment, 2) the tarped unamended control with each tarped treatment, and 3) the tarped with the untarped treatments for each amendment that was treated both ways. Bonferroni’s test (P=0.05) was used for the mean separation of these contrasts. For the cover cropping experiment, average disease ratings per pot were used for 2-way ANOVA testing of treatments, greenhouse location, and treatment and treatment interactions, using SigmaStat (Jandel Scientific Software, San Rafael, CA). Dunnett’s test (P=0.05) was used for mean separation of each treatment with the fallow control. FOA2 isolations from celery crowns Florida 683 celery crowns with different disease ratings were collected from the amendment experiments. Crowns were rinsed in water and washed three times for 30 84 minutes in 1% sodium hexametaphosphate. The crowns were then washed in 60% ethanol (1 minute) and 0.5 % sodium hypochlorite with 2 drops of Tween 20 (5 minutes) followed by 3 rinses in sterile water. Pieces of celery crown were extracted from the crown interior and plated on Komada’s medium. Mycelium growing from these crowns were transferred to PDA. Identification of the isolates as FOA2 was done with colony size on D-medium (Puhalla 1984). Twenty stab transfers were made from the colony on PDA onto D-medium in a procedure described by Correll et al. (1986) and Jacobson (unpublished). Controls used were Michigan isolates known to be FOA2. Colony diameters were measured under a dissecting microscope equipped with an ocular micrometer, averaged for each isolate, and relative colony diameters were produced by dividing the average diameter of the unknown isolate by the average diameter of the controls. Relative colony diameters of 1.2 or less are highly likely to be FOA2 (D.J. Jacobson, unpublished). 85 RESULTS Determination of soil infestation method Naturally infested VerHage muck soil with additional artificial infestation supported the highest disease ratings, followed by artificially infested Baccto" potting mix and artificially infested muck soil from the MSU Muck farm which were not naturally infested with FOA2 (Figure 2). Celery fresh weight was raised substantially in potting mix compared with other soils tested (Figure 2). The only treatment combination that was significantly higher (P=0.05) than the naturally infested control was the use of naturally infested VerHage soil supplemented with 3.08 g/kg of wheat straw colonized by the FOA2 isolate from Hudsonville, MI (Figure 3). Although disease was not significantly affected by depth of planting when tested statistically (P=0.05), deep planting of seedlings tended to increase disease levels in naturally infested soils (Figure 3). Doubling the inoculum concentration also did not increase disease (Figure 3). Adding colonized wheat straw to steamed MSU muck soil did not substantially increase Fusarium yellows compared with artificially infested unsteamed MSU muck soil unless the inoculum was not dried (Figure 4). Steamed soil infested with undried inoculum raised disease levels (Figure 4), although the mean disease rating was lower than with naturally infested VerHage soil supplemented with additional FOA2-infested straw (Figures 2 and 3) (P=0.05). Steamed muck soil produced celery with higher fresh weight than muck soil that was not steamed (Figure 4). 86 Figure 2. Effect of soil type on mean Fusarium yellows disease ratings and fresh weights of Florida 683 celery seedlings 6 weeks after transplanting into soil artificially infested with low levels of FOA2. lDisease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2Mean fresh weight (g) of celery per pot. 3Three soil types examined were: VerHage soil = FOA2 naturally infested muck soil, MSU soil = FOA2-free muck soil, and Potting mix = Baccto‘2 Potting Mix. For each treatment, soil was infested with FOA2 at a rate of 3.08g/kg soil. ‘Four-week old Florida 683 celery seedlings were planted with the crowns 1-2 cm below the soil line. 5Four-week old Florida 683 celery seedlings were planted with the crown at the soil line. Mean Disease Rating1 Mean Fresh Weight (g)2 87 E: Deep Planting‘ sauna. Homing: s\\\\\\\\F-+ //, He; 50 Naturally infested ArtIfICIOIIy mfested Art1f1c1ally1nfested Artificmlly Infested Potting Mix VerHage Soil VerHage Soil Treatment3'M 40- 30- 20- 10- : Deep Planting‘ Shallow Plantin95 km 1\\\\\ Q / s\\\\\\\\\\\\\\\\\8 Naturally infested Artificially infested VerHage Soil VerHage eSoil r Artificially infested Art1f1C1ally infested Potting Mix MSU Soil eatment3 88 Figure 3. Effect of planting depth, inoculum level, and isolate type on the mean Fusarium yellows disease ratings and fresh weights of Florida 683 celery seedlings, 6 weeks after transplanting, using naturally infested soil from VerHage farms, Hudsonville, MI. 1Disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2Mean fresh weight (g) of celery per pot. 3Treatment designations are as follows: control = naturally infested VerHage soil only low = 3.08 g colonized wheat straw inoculum/kg soil high = 6.16 g colonized wheat straw inoculum/kg soil Whitehall isolate = FOA2 isolate from Whitehall, MI Hudsonville isolate = FOA2 isolate from Hudsonville, MI ‘Four-week old Florida 683 celery seedlings were planted with the crowns 1-2 cm below the soil line. SFour-week old Florida 683 celery seedlings were planted with the crowns at the soil line. 89 TW/ 2 Deep Planting4 EZZZI Shallow Plant1ng5 /////////A 1% 1% 1-1111 11"I High Whitehall isolate Low . Hudsonw Treatment-3 bbbbbbb Foczom 3035 c002 08642086420 111111 «A3 E903 smut c022 90 Figure 4. Effects of steaming MSU muck soil and using wet or dry FOA2-colonized wheat straw inoculum on mean Fusarium yellows disease ratings and fresh weights of Florida 683 celery seedlings, 6 weeks after transplanting. lDisease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2Mean fresh weight (g) of celery per pot. 3FOA2-colonized wheat straw was either dried prior to incorporation with MSU muck soil or incorporated directly into MSU muck soil without drying at rates of 3.08g/kg soil. 91 . I: Unsteamed Soil l— Steamed Soil Mean Disease Rating1 01 Wet ’ .. 07% , 1 %. 2 Wet Dry Treatment~'5 0' O E Unsteamed Sail a O N O Mean Fresh Weight (g)2 u o _n O Dry Wet Dry Wet Treatment3 92 No strong correlations existed between disease rating and fresh weight. The overall correlation coefficient (excluding steamed soil treatments) was —0.22. Correlations between disease rating and weight for each soil type were 0.01, 0.35, and -0.44 for VerHage muck soil, potting mix, and MSU muck soil, respectively. These results indicated that supplementing naturally FOA2-infested muck soil from VerHage Farms, Hudsonville, MI with wheat straw colonized by the FOA2 isolate from Hudsonville, M1 at a rate of 3.08g/kg would produce the highest level of Fusarium yellows of celery if one-month-old Florida 683 seedlings were transplanted with the crowns 2-3 cm below the soil surface. Effect of organic amendments on Fusarium yellows of celery Most soil amendments in artificially infested VerHage soil screened without tarping or heating did not affect disease severity compared with the infested fallow control in the first study, except that the celery amendment significantly increased disease at P=0.05 (Figure 5A). In the second study, no differences existed compared with the unamended infested fallow control (Figure SB). Disease rating in the infested fallow control was significantly different from the noninfested control for both studies (Figure 5). Greenhouse location did not affect disease ratings and there was no treatment X location interaction. In the first study, differences were observed between the tarped and heated infested fallow control and other tarped and heated treatments. Tarping and heating soil did not lower disease of the infested control (Figure 5A). Tarping and heating mustard- amended soil appeared to lower disease severity compared with the tarped infested fallow 93 Figure 5. Effect of soil amendments on mean Fusarium yellows disease ratings of Florida 683 celery seedlings 8 weeks after transplanting into FOA2-infested soil in the greenhouse. Results from two studies are provided separately (A, B). SEM (standard error of the means) for A = 0.159; SEM for B = 0.173 1Mean disease ratings of 15 replicates were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2Fresh amendments were incorporated into soil at a rate of 2% (w/w). Treatments of cabbage, mustard, onion, unamended control (fallow), and an noninfested control (no FOA2) were also enclosed in plastic and heated under lamps prior to celery transplanting. 1 Mean Disease Rating 1 Mean Disease Rating 94 [:3 Without tarping and heating A - With tarping and heating (celery, rape, and rye not tested) 3 _ i— F— i— r— 2 _ (1) 1 no FOA2 fallow cabbage mustard onion celery rape rye Treatment2 5 1:1 VVrthout tarping and heating B 4 _ - With tarping and heating (celery, rape, and rye not tested) 3 .— __ 7— _ 2 __ m j i I no FOA2 fallow cabbage mustard onion celery rape rye Tre atrnent2 95 control but the difference was not significant (Figure 5A). The tarped onion amendment treatment lowered disease severity of Fusarium yellows as low as the noninfested controls (P=0.01) (Figure 5A). Tarping and heating treatments in the second study (Figure 5B) were not statistically different (P=0.05) from the noninfested control. Although statistically insignificant, there was a decrease in disease pressure in the unamended infested fallow control following tarping and heating during the second study (Figure 5B). In both studies, mustard and onion amendments significantly lowered disease severity when the soil was tarped and heated, compared to treatments in which the pots were left open after amending (Figure 5A,B). Tarping cabbage-amended soils significantly (P=0.01) lowered Fusarium yellows in one of the two studies. Effect of cover crops on Fusarium yellows of celery None of the cover crops screened affected disease severity of Fusarium yellows of celery when compared with the unamended infested fallow control in either study (Figure 6A,B). Disease pressure in the second study was very low; the infested fallow control had a mean disease rating of 1.8 which was not significantly different from the noninfested control (Figure 6B). Greenhouse location did not affect disease ratings and there was no significant interaction between treatment and location. FOA2 isolations from celery crowns Fusarium oxysporum was isolated from Florida 683 celery crowns with disease ratings of 1—4 (Table 8). Isolation of Fusarium oxysporum from celery plants with a 96 Figure 6. Effect of cover crops on mean Fusarium yellows disease ratings of Florida 683 celery seedlings 8 weeks after transplanting into FOA2-infested soil in the greenhouse. Results from two studies are provided separately (A, B). SEM (standard error of the means) for A = 0.232; SEM for B = 0.211 lMean disease ratings were based on a 1 to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown area; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored; 5 = dead or nearly dead plant. 2Cover crops were grown in pots for 2 weeks and incorporated into soil. A fallow treatment and steamed soil (no FOA2) were controls. 1 Mean Disease Rating 1 Mean Disease Rating 97 5 A 4 _ 3 _. 2 _. (1) 1 no FOA2 fallow cabbage mustard celery rape rye Treatment2 5 B 4 __ 3 .— 2 fl (1) 1 no FOA2 fallow cabbage mustard celery rape rye Treatment2 98 disease rating of 5 (dead or nearly dead) was unsuccessful because of the contamination from secondary microorganisms. Fusarium oxysporum f. sp. apii race 2 has been identified as colonies grown on D-medium with a relative average diameter less than or equal to 1.0 compared to known FOA2 isolates (Correll et al. 1986). Relative average diameters of 1.2 or less are also usually FOA2 (D.J. Jacobson, unpublished). Although some FOA2 isolates have had relative diameters greater than 1.2, most Fusarium oxysporum with relative colony diameters above 1.2 are non-pathogenic on celery. Relative colony diameters greater than 1.2, therefore, are unlikely to be FOA2. All isolates from infected crowns with a rating of 3 or 4 had relative colony diameters on D- medium less than 1.2 (Table 8). All these isolates are likely to be FOA2. Sixty percent of the isolates from crowns with a disease rating of 2 had relative colony diameters less than 1.2 (Table 8). From celery crowns that did not have any vascular discoloration, 45% of the isolates had relative colony diameters less than 1.2, and they were likely to be FOA2 (Table 8). 99 Table 8. Percentage of Fusarium oxysporum f. sp. apii race 2 isolates compared with total Fusarium oxysporum isolated from Florida 683 celery crowns, based on colony size on the selective D-medium. Percent of isolates with relative colony diameter (RCD)2 Disease Ratingl RCD s 1.2 RCD > 1.2 1 (n=27) 45% 55% 2 (n=5) 60% 40% 3 (11:5) 100% 0% ,4 (n=5) 100% 0% 1 ‘Disease ratings were based on a l to 5 scale, where: 1 = no vascular discoloration; 2 = trace of vascular discoloration in root or crown; 3 = < 50% crown area discolored; 4 = 2 50% crown area discolored. 2Relative colony diameter was measured by dividing the average diameter of 20 colonies at 72 hours on D-medium by the average diameter of the 20 control (known FOA2) colonies. Colonies with an average relative colony diameter less than 1.2 are usually identifiable as FOA2 (D.J. Jacobson, unpublished). 100 DISCUSSION Greenhouses offer controlled environments in which to screen soil amendments for their effectiveness in soilbome disease control. The value of greenhouse studies has been questionable for disease control experiments because often the results cannot be repeated in the field (Patrick and Toussoun 1970, Zakaria and Lockwood 1980). These discrepancies have produced "feelings of frustration and disappointment in attempts to make practical use of biological control" (Lewis and Papavizas 1975). Greenhouse conditions attempt to remove sources of variability, strengthening statistical evidence for treatment differences observed. Although the unnatural greenhouse environment may show differences between treatments, once in the field the naturally occurring variability may be large enough to render the differences insignificant. However, greenhouses can provide valuable information about the potential success of soil amendments for soilbome disease control. Whereas a successful greenhouse control must be tested for field success, treatments that do not work in the greenhouse are generally not worth pursuing in the field. Although greenhouses provide well-controlled environments, they may be difficult to use because of an inability to replicate conditions necessary to produce infection, as was seen in this study. Fusarium yellows of celery is a disease for which the environmental factors regulating infection and disease are not well understood (Toth 1989, Hart and Endo 1981). Initially, low disease levels and highly variable disease reactions hindered this study by preventing any potential differences from being observed or separated. Maintaining soil temperatures at 22-24°C, reducing fertilizer applications, 101 withholding water, saturating soil with water, adding FOA2 chlamydospores to soil, dipping celery roots into a FOA2 chlamydospore suspension, or mixing a FOA2- colonized agar plate into the soil did not raise disease ratings. Elmer (1985) had successful expression of Fusarium yellows of celery by infesting soil with FOA2- colonized wheat straw. The best variables among inoculum level, soil type, isolate, and wetness of the inoculum were determined for use with Elmer’s (1985) soil infestation technique. The best conditions for disease expression were produced by adding 3.08 g of wheat straw colonized by the Hudsonville isolate to 1 kg of naturally infested muck soil, and planting one-month old Florida 683 seedlings with their crowns 2-3 cm below the soil line. Even so, the average disease rating was only between 2 and 3 (Figures 5 and 6). Fluctuating day and night temperatures as well as higher temperatures and light intensity also improved disease severity of Fusarium yellows of celery somewhat (K.V. Subbarao personal communication, Toth 1989). The amendments and cover crops screened were chosen partly because they provided control of other soilbome Fusarium species. Rye is commonly used as a cover crop by celery growers to reduce erosion of organic soils (Toth 1989). It has also helped control root rot of bean caused by Fusarium solani f. sp. phaseoli if amended into soil and given ample time to decompose (Lewis and Papavizas 1975). Cabbage has been successfully used to control soilbome fungi (Lewis and Papavizas 1971, Ramirez- Villapudua and Munnecke 1987), possibly because of the release of toxic volatiles. Other crucifers chosen were mustard and rape because they have higher levels of glucosinolates than many other members of the Brassicacea and they are widely available (Carlson et al. 1987). Glucosinolates liberated isothiocyanates and thiocyanates during 102 decomposition which are biological toxins (Carlson et al. 1987, Lewis and Papavizas 1971). Compared with mustard, rape also grew more foliage so that the amount of amendment reincorporated after cover cropping was expected to be much higher. The amount of amendment has been shown to be proportional to the effectiveness of control (Zakaria and Lockwood 1980). Allium spp. have been useful for field control of FOA2 (T 0th 1989). High sugar content in onions is thought to stimulate chlamydospore germination which may help to control disease (Elmer 1985). Onion also produced alkyl thiosulphinates which have antibiotic properties to some Fusarium species (Agrawal 1978). Onion residues inhibited disease development when amended into soils at very high rates (Ehner and Lacy 1987a). Celery was used as a check, since incorporating celery into FOA2-infested soils increased Fusarium yellows of celery (Ehner and Lacy 1987a). None of the crops tested significantly lowered the levels of Fusarium yellows in celery when used as soil amendments or cover crops. The only statistical difference from the fallow control was with celery, which increased disease slightly in one of the amendment experiments. Soil amendments may have supported the growth of FOA2 or the formation of daughter chlamydospores of FOA2. Elmer and Lacy (1987a) showed that population levels of onion and mint were similar to a fallow control 2-3 weeks after incorporation, which may account for the lack of differences seen in the amendment experiments. Celery residues maintained higher population levels than the fallow during the first 3 weeks after soil amending (Elmer and Lacy 1987a) which may be the reason celery residues increased disease levels (Figure 5A). Amendments were incorporated at levels of 2% (w/w) which would be equivalent to 10-20 tons/acre-furrow slice (Johnson 103 and Curl 1972). To control root diseases, amendments must usually be incorporated at level at least this high (Johnson and Curl 1972, Papavizas and Davey 1960, Zakaria and Lockwood 1980) which is considered to be irnpractically high for field use (Johnson and Curl 1972, Papavizas 1975). Crop residues were the most significant aspect in crop rotation for the control of soilbome fungi (Williams and Schmitthenner 1962). However, residues left in the field following crop rotation are from mature plants and have a high C:N ratio. Although C:N ratio does not always correlate with disease control (Huber and Watson 1970), soilbome Fusarium diseases have been more readily controlled with mature (high C:N) residues than immature (low C:N) ones (Papavizas et al. 1968). All of the residues used to screen for Fusarium yellows of celery control were young and were likely to have a low C:N ratio. Possibly, this may account for the poor results. Cover crops differed from amendment experiments by exposing inoculum to non- host roots. Root exudates tended to stimulate chlamydospores to break dormancy (Lockwood 1988), allowing roots to be colonized by non-pathogens, such as FOA2 (Elmer 1985, Elmer and Lacy 1987a). If cover crop roots were colonized by FOA2, they may have been able to survive as hyphae in roots or formed daughter chlamydospores during the decomposition of the incorporated cover crops. The cover cropping experiments had different average levels of disease expression (Figure 6). Mean disease ratings were so low in the second experiment that they were insignificantly greater than the uninfested control. Since the second experiment was set up after the greenhouse had been shaded in order to lower temperatures, lower temperatures and 104 lower light intensity may be partially responsible for the reduced disease levels observed (T 0th 1989, K.V. Subbarao, personal communication). It is difficult to assess the impact of greenhouse disease rating on celery productivity. The growing conditions under which the studies were conducted were very different than growing conditions in the field which may have affected disease severity. Celery plants were transplanted at 4 weeks old and rated at 8 weeks following transplanting. In the field, plants are transplanted at 7 weeks and harvested 3 months after transplanting. As well, the seedlings in 10 cm pots have restricted growth and are usually root-bound within 4 or 5 weeks. Since FOA2 enters through the root tips (Hart and Endo 1981, Schneider 1984), very little root infection likely occurred once the plants were root-bound. By this time, there may not have been sufficient numbers of infection to produce disease (Hart and Endo 1981). However, in the greenhouse studies, the amount of inoculum was unnaturally high at expected levels of 100,000 colony forming units per gram of soil above the natural infestation level (Elmer 1985) so that high levels of infection were likely if conditions for infection were suitable. It was interesting that colonies likely to be FOA2 were isolated from celery crowns that did not show any symptoms of infection (Table 8). Since this occurred in nearly half the plants tested, it was unlikely to be surface contamination. Since isolations were made from crown tissue with the outer cortical cells removed, it appeared that the FOA2 isolates had colonized or grown into the vascular tissue. In a few cases, Fusarium species likely to be FOA2 were isolated from petioles of celery with no vascular discoloration in the crown. Insufficient time or improper environmental conditions may have prevented symptom expression. Perhaps if the experiments had lasted for a longer 105 time, disease expression would have occurred and disease ratings would have been higher. It is not certain how infection (as opposed to disease symptoms) affects celery yield and marketability. Elmer and Lacy (1987a) found that celery crown were colonized by FOA2 regardless of the resistance level of the cultivar. Although infection occurred in all cultivars tested, phenolic compounds slowed disease development in resistant cultivars (Jordan et al. 1989). Since only the highly susceptible cultivar Florida 683 was used for this study, it was likely that all infected crowns would eventually deve10p vascular discoloration. Identification of FOA2 using only the criterion of colony size on D-medium may not be sufficient for any isolates with average relative diameters greater than 1.0 or 1.2 (D.J . Jacobson, unpublished). Pairing nitrate non-utilizing (nit) mutants of the isolates with known FOA2 nit mutants would have given stronger evidence for the identification of FOA2 (Toth 1989). Unfortunately, one of the tester nit mutants used tended to revert back to wild type growth, preventing accurate identification of FOA2 colonies using this technique. Although the soil amendment experiments did not show disease control promise, tarping and heating for 2 weeks after soil amending did produce very satisfactory disease control for some treatments. In both studies, tarping and heating of mustard- and onion- amended soils lowered disease levels significantly when compared to mustard- and onion- amended soils without tarping or heating. In the first experiment, tarping and heating did not affect disease levels in unamended soil or soil amended with cabbage tissue. The significant drop in disease levels of all tarped and heated treatments in the second study (Figure SB) may be attributed to toxic compounds released by the plastic pots used. In 106 the second study, new plastic pots were used that were not used during the first study. These pots were deformed during the heating period and a very strong plastic smell was noticeable as the tarps were removed. During heating, these pots may have released fungitoxic or fungistatic compounds that accounted for the low disease observed in all treatments. It is uncertain whether tarping and heating of cabbage-amended or unamended soil has a depressive effect on Fusarium yellows of celery because of the conflicting results of the studies. Since mustard and onion did not provide Fusarium yellows control in the greenhouse (Figures 5 and 6) or field (Toth 1989) unless tarped, it was likely that mustard and onion produced volatile toxins that affected FOA2 survival. Glucosinolates and isothiocyanates are released from crucifers (Carlson et a1. 1987, Lewis and Papavizas 1970) and alkyl thiosulphinates from onions (Agrawal 1978) may affect viability of FOA2 propagules. It was surprising that cabbage amendments did not consistently provide disease control when used with tarping and heating. Cabbage has effectively controlled F usarium oxysporum f. sp. conglutinans (Ramirez-Villapudua and Munnecke 1987) and produces high levels of fungitoxic compounds similar to other crucifers (Lewis and Papavizas 1970). However, different crucifers can have dramatically different effects on the disease control (K.V. Subbarao, personal communication). The effect on disease with tarping and heating may be two-fold. Although toxic volatiles are likely to affect the pathogen directly (Lewis and Papavizas 1970) by weakening or killing the pathogen, the release of volatile compounds and heating may encourage the growth of heat-resistant soilbome antagonists that may lyse or compete 107 with the pathogen and provide disease control (Gamliel and Katan 1993, Katan 1985, Katan et al. 1976, Ramirez-Villapudua and Munnecke 1986). The selective stimulation of antagonistic bacteria by certain amendments may partially account for the poor disease control provided by cabbage in the first study. Based on the results from these greenhouse studies, none of the crops surveyed can be recommended as cover crops or soil amendments in the field for disease control. Mustard or onion residues could have potential field use if they are tarped. Unfortunately, in Michigan it is unlikely that conditions would be suitable for tarping to achieve sufficiently high soil temperatures for extended periods (Katan 1981). It would be important to determine if disease would be reduced by tarping soil amended with mustard or onion without heating. Also, dried amendments may improve disease control since they decompose faster than fresh amendments (Ramirez-Villapudua and Munnecke 1988). These studies only examined disease control over short periods of time and at unnaturally high inoculum levels. Field evidence exists that Allium spp. may provide good Fusarium disease control if used in rotation (Granges 1992, R. Schreur, personal communication, Toth 1989). A one year leek rotation significantly reduced Fusarium populations (Toth 1989) and a four year rotation into onions allowed a heavily FOA2- infested muck farm to again be used for successful celery production (R. Schreur, personal communication). LITERATURE CITED Adams, P.B., J .A. Lewis, and G.C. Papavizas. 1968a. Survival of root-infecting fungi in soil. III. The effect of cellulose amendment on chlamydospore germination of Fusarium solani f.sp. phaseoli. Phytopathol. 58:373-377. Adams, P.B., J .A. Lewis, and G.C. Papavizas. 1968b. Survival of root-infecting fungi in soil. IV. The nature of fungistasis in natural and cellulose-amended soil on chlamydospores of Fusarium solani f. sp. phaseoli. Phytopathol. 58:378-383. Agrawal, P. 1978. Effect of root and bulb extracts of Allium spp. on fungal growth. Trans. Br. Mycol. Soc. 70:439-441. Agrios, G.N. 1988. Plant Pathology 3rd edition. Academic Press, Inc. San Diego, CA. p.803. Awuah, R.T. and J.W. Lorbeer. 1991. Methyl bromide and steam treatment of an organic soil for control of Fusarium yellows of celery. Plant Dis. 75:123-125. Awuah, R.T., J .W. Lorbeer, and L.A. Ellerbrock. 1986. Occurrence of Fusarium yellows of celery caused by Fusarium oxysporum f.sp. apii race 2 in New York and its control. Plant Dis. 70:1154-1158. Becker, J.O., C.A. Hepfer, G.Y. Yuen, S.D. Van Gundy, M.N. Schroth, J.G. Hancock, A.R. Weinhold, and T. Bowman. 1990. Effect of rhizobacteria and metham-sodium growth and root microflora of celery cultivars. Phytopathol. 80:206- 211. Burpee, LL. 1990. The influence of abiotic factors on biological control of soilbome plant pathogenic fungi. Can. J. Plant Pathol. 12:308-317. Carlson, D.G., M.E. Daxenbichler, C.H. VanEtten, W.F. Kwolek, and PH. Williams. 1987. Glucosinolates in crucifer vegetables: broccoli, brussels sprouts, cauliflower, collards, kale, mustard greens, and kohlrabi. J. Am. Soc. Hort. Sci. 112:173-178. 108 109 Cook, R.J. and MN. Schroth. 1965. Carbon and nitrogen compounds and germination of chlamydospores of Fusarium solani f. sp. phaseoli. Phytopathol. 55:254-256. Elmer, W.H. 1985. The ecology and control of Fusarium yellows of celery in Michigan. PhD. Dissertation. Michigan State University, East Lansing, MI. p.146. Elmer, W.H. and M.L. Lacy. 1987a. Effects of crop residues of colonization of plant tissues on propagule survival and soil populations of Fusarium oxysporum f.sp. apii race 2. Phytopathol. 77:381-387. Elmer, W.H. and M.L. Lacy. 1987b. Effects of inoculum densities of Fusarium oxysporum f. sp. apii in organic soil on disease expression in celery. Plant Dis. 71 : 1086-1089. Ford, E.J., A.H. Gold, and W.C. Snyder. 1970. Induction of chlamydospore formation in Fusarium solani by soil bacteria. Phytopathol. 60:479-484. Gamliel, A. and J. Katan. 1993. Suppression of major and minor pathogens by fluorescent pseudomonads in solarized and nonsolarized soils. Phytopathol. 83:68-75. Garrett, SD. 1970. Pathogenic Root-Infecting Fungi. Cambridge University Press, Cambridge, MA. p.294. Gill, IL. 1978. Design and analysis of experiments. Iowa State University Press, Ames, Iowa. p.411. Gilligan, CA. 1986. Use and misuse of the analysis of variance in plant pathology. Adv. Plant Pathol. 5:225-261. Granges, A. 1992. La rotation des cultures maraicheres. Revue Suisse Vitic. Arboric. Hortic. 24:21-24. Griffin, G.J. 1981. Physiology of conidium and chlamydospore germination in Fusarium. pp. 331-339. In: Fusarium: Diseases, Biology, and Taxonomy. eds. Nelson, P.E., T.A. Toussoun, and R.J. Cook. Pennsylvania University Press, University Park, PN. p.457. Ham, J.M., G.J. Kluitenberg, and W.J. Lamont. 1993. Optical properties of plastic mulches affect the field temperature regime. J. Am. Soc. Hort. Sci. 118:188-193. 110 Hardy, G.E.S.J and K. Sivasithamparam. 1985. Soil solarization: effects on Fusarium wilt of carnation and Verticillium wilt of eggplant. pp.279-281. In: Ecology and Management of Soilborne Plant Pathogens: Proceedings of Section 5 of the Fourth International Congress of Plant Pathology. eds. Parker, C.A., A.D. Rovira, K.J. Moore, and P.T.W. Wong. APS Press, St. Paul, MN. p.358. Hart, LP. and R.M. Endo. 1981. The effect of time of exposure to inoculum, plant age, root development, and root wounding on Fusarium yellows of celery. Phytopathol. 71:77-79. Hoitink, H.A.J. and RC. Fahy. 1986. Basis for the control of soilbome plant pathogens with composts. Ann. Rev. Phytopath. 24:93-114. Huber, D.M. and RD. Watson. 1970. Effect of organic amendment on soil-borne plant pathogens. Phytopathol. 60:22-26. Jarvis, W.R. and HI. Thorpe. 1981. Control of fusarium foot and root rot of tomato by soil amendment with lettuce residues. Can. J. Plant Pathol. 3:159-162. Johnson, KB. 1994. Dose-response relationships and inundative biological control. Phytopathol. 84:780-784. Johnson, LP. and E.A. Curl. 1972. Methods for Research on the Ecology of Soil- Bome Plant Pathogens. Burgess Publishing Co. Minneapolis, MN. p. 247. Katan, J. 1981. Solar heating (solarization) of soil for control of soilbome pests. Ann. Rev. Phytopath. 19:211-236. Katan, J. 1985. Solar disinfestation of soils. pp. 274-277. In: Ecology and Management of Soilborne Plant Pathogens: Proceedings of section 5 of the fourth International Congress of Plant Pathology. eds. Parker, C.A., A.D. Rovira, K.J. Moore, and P.T.W. Wong. APS Press, St.Paul, MN. p.358. Katan, J ., A. Greenberger, H. Alon, and A. Grinstein. 1976. Solar heating by polyethylene mulching for the control of diseases caused by soil-borne pathogens. Phytopathol. 66:683-688. Lewis, J .A. and G.C. Papavizas. 1970. Evolution of volatile sulfur-containing compounds from decomposition of crucifers in soil. Soil Biol. Biochem. 2:239-246. Lewis, J.A. and G.C. Papavizas. 1971. Effect of sulfur-containing volatile compounds and vapors from cabbage decomposition on Aphanomyces euteiches. Phytopathol. 61 :208-214. 111 Lewis, J .A. and G.C. Papavizas. 1975. Survival and multiplication of soil-borne plant pathogens as affected by plant tissue amendments. pp. 84-89. In: Biology and Control of Soil—Bome Plant Pathogens. ed. Bruehl, G.W. p.216. Lewis, J.A. and G.C. Papavizas. 1977. Effect of plant residues on chlamydospore germination of Fusarium solani f. sp. phaseoli and on Fusarium root rot of bean. Phytopathol. 67:925-929. Linderrnan, R.G. 1970. Plant residue decomposition products and their effects on host roots and fungi pathogenic to roots. Phytopathol. 60: 19-22. Lockwood, J .L. 1988. Evolution of concepts associated with soilbome plant pathogens. Ann. Rev. Phytopath. 26:93-121. Lumsden, R.D., J .A. Lewis, and RD. Millner. 1983. Effect of composted sewage sludge on several soilbome pathogens and diseases. Phytopathol. 73:1543-1548. Menzies, J .D. and R.G. Gilbert. 1967. Responses of the soil microflora to volatile components in plant residues. Soil Sci. Soc. Am. Proc. 31:495-496. Mitchell, R. and M. Alexander. 1961a. Chitin and the biological control of Fusarium diseases. Plant Dis. Rep. 45:487-490. Mitchell, R. and M. Alexander. 1961b. The mycolytic phenomenon and biological control of Fusarium in soil. Nature 190:109-110. Nash, SM. and W.C. Snyder. 1962. Quantitative estimations by plate counts of propagules of the bean root rot Fusarium in field soils. Phytopathol. 52:567-572. Nelson, P.E., T.A. Toussoun, and R.J. Cook. 1981. Fusarium: Diseases, Biology, and Taxonomy. Pennsylvania State University Press, University Park, Pennsylvania. p. 193. Opgenorth, D.C. and R.M. Endo. 1983. Evidence that antagonistic bacteria suppress Fusarium wilt of celery in neutral and alkaline soils. Phytopathol. 73:703- 708. Opgenorth, D.C. and R.M. Endo. 1985. Abiotic factors and chlamydospore formation in Fusarium oxysporum f. sp. apii. Trans. Brit. Mycol. Soc. 84:740-742. Oritsejafor, J .J . and M.O. Adeniji. 1990. Influence of host and non-host rhizospheres and organic amendments on survival of Fusarium oxysporum f. sp. elaeidis. Mycol. Res. 94:57-63. 112 Otto, H.W., A.O. Paulus, M.J. Snyder, R.M. Endo, L.P. Hart, and J. Nelson. 1976. A crown rot of celery. Cal. Agr. 30:10-11. Papavizas, G.C. and CB. Davey. 1960. Rhizoctonia disease of bean as affected by decomposing green plant materials and associated microfloras. Phytopathol. 50:516- 522. Papavizas, G.C., J.A. Lewis, and PB. Adams. 1968. Survival of root-infecting fungi in soil. 11. Influence of amendment and soil carbon-to-nitrogen balance on Fusarium root rot of bean. Phytopathol. 58:365-372. Papavizas, G.C. and RD. Lumsden. 1980. Biological control of soilborne fungal propagules. Ann. Rev. Phytopath. 18:389—413. Park, D. 1954. Chlamydospores and survival in soil fungi. Nature 173:454-455. Patrick, Z.A. and T.A. Toussoun. 1970. Plant residues and organic amendments in relation to biological control. pp. 440-459. In: Ecology of Soil-Borne Plant Pathogens: Prelude to Biolcygical Cm. eds. K.F. Baker and W.C. Snyder. University of California Press, Berkeley, CA. p.571. Price, D. 1984. Fusarium and plant pathology: the reservoir of infection. pp.71-93. In: The Applied Mycology of Fusarium. eds. Moss, M.O. and J .E. Smith. Cambridge University Press, Cambridge. p.264. Ramirez-Villapudua, J. and DE. Munnecke. 1986. Solar heating and amendments control cabbage yellows. Cal. Agr. 40:11-13. Ramirez-Villapudua, J. and DE. Munnecke. 1987. Control of cabbage yellows (Fusarium oxysporum f. sp. conglutinans) by solar heating of field soils amended with dry cabbage residues. Plant Dis. 71:217-221. Ramirez-Villapudua, J. and DE. Munnecke. 1988. Effect of solar heating and soil amendments of cruciferous residues on Fusarium oxysporum f. sp. conglutinans and other organisms. Phytopathol. 78:289-295. Reyes, AA. and J .E. Mitchell. 1962. Growth response of several isolates of Fusarium ion rhizospheres of host and nonhost plants. Phytopathol. 52:1196-1200. Schippers, B. and W.H. van Eck. 1981. Formation and survival of chlamydospores in Fusarium. pp. 250-260. In: Fusarium: Diseases, Biology, Taxonomy. Pennsylvania State University Press, University Park, PN. p.457. 113 Schneider, R.W. 1984. Effects of nonpathogenic strains of Fusarium oxysporum on celery root infection of Fusarium oxysporum f. sp. apii and a novel use of the Lineweaver-Burk double reciprocal plot technique. Phytopathol. 742646-653. Schneider, R.W. 1985. Suppression of Fusarium yellows of celery with potassium, chloride, and nitrate. Phytopathol. 75:40-48. Schroth, MN. and JG. Hancock. 1982. Disease-suppressive soil and root- colonizing bacteria. Science 216:1376-1381. Schroth, MN. and F.F. Hendrix, Jr. 1962. Influence of nonsusceptible plants on the survival of Fusarium solani f.sp. phaseoli in soil. Phytopathol. 52:906-909. Schroth, M.N . and D.C. Hildebrand. 1964. Influence of plant exudates on root- infecting fungi. Ann. Rev. Phytopathol. 22101-133. Sequeira, L. 1962. Influence of organic amendments on survival of Fusarium oxysporum f. cubense in the soil. Phytopathol. 52:976-982. Sewell, G.W.F. 1970. The effect of altered physical condition of soil on biological control. pp.479-494. In: Ecologv of Soil-Bome Plant Pathogens: Prelude to Biological Control. Eds. K.F. Baker and W.C. Snyder. University of California Press, Berkeley, CA. p.571. Smith, SN. and W.C. Snyder. 1972. Germination of Fusarium oxysporum chlamydospores in soils favorable and unfavorable to wilt establishment. Phytopathol. 62:273-277. Sneh, B., M. Dupler, Y. Elad, and R. Baker. 1984. Chlamydospore Germination of F usarium oxysporum f. sp. cucumerinum as affected by fluorescent and lytic bacteria from a Fusarium-suppressive soil. Phytopathol. 74:1115-1124. Sun, S. and J. Huang. 1985. Formulated soil amendment for controlling Fusarium wilt and other soilborne diseases. Plant Dis. 69:917-920. Toth, K.F. 1989. Biology and control of Fusarium oxysporum f.sp. apii race 2. PhD. Dissertation. Michigan State University, East Lansing, MI. p. 174. Toth, K.F. and M.L. Lacy. 1991. Increasing resistance in celery to Fusarium oxysporum f. sp. apii race 2 with somaclonal variation. Plant Dis. 75:1034-1037. Toussoun, T.A., W. Menzinger, and RS. Smith Jr. 1969. Role of conifer litter in ecology of Fusarium: stimulation of germination in soil. Phytopathol. 59:1396-1399. 114 Williams, LE. and AF. Schmitthenner. 1962. Effect of crop rotation on soil fungus populations. Phytopathol. 52:241-246. Zakaria, M.A. 1978. Reduction in Fusarium populations in soil by oilseed meal amendments. PhD Dissertation. Michigan State University, East Lansing. p.86. Zakaria, M.A. and J .L. Lockwood. 1980. Reduction in Fusarium populations in soil by oilseed meal amendments. Phytopathol. 70:240-243. CHAPTER 4 - EFFECT OF FURANOCOUMARIN ON FUSARIUM OXYSPORUM F .SP. APII RACE 2 INTRODUCTION Furanocoumarins (syn. furocoumarins) are compounds that are best known for their ability to incite photoirritant contact dermatitis (syn. photodermatitis) (DeL.eo 1992). Furanocoumarin-containing items that have commonly induced these skin disorders are perfume, limes, and celery (Kagan 1993). Dermatological effects following furanocoumarin exposure commonly include the formation of rash or blister-like bullous lesions and hyperpigmentation (Berkley et al. 1986, DeLeo 1992). In plants, furanocoumarins have provided chemical defense against attack by microorganisms (Beier and Oertli 1983, Johnson et al. 1973) and insect pests (Berenbaum 1981, Trumble et al. 1990). Furanocoumarins are naturally occurring compounds that are most commonly produced in members of the Apiaceae and Rutaceae families, although they have also been found in plants from the Moraceae, Asteraceae, Fabaceae, and Rosaceae (Kagan 1993). Plant extracts that contain photoactive furanocoumarins have been used for at least 3000 years to treat vitiligo, a skin condition in which skin pigmentation is lost due to melanocyte malfunction (Scott et al. 1976). Furanocoumarins are now used clinically 115 116 with UV exposure to stimulate melanin production in pigmentless skin of vitiligo patients (Scott et al. 1976). Another common clinical use for furanocoumarins is the treatment of psoriasis. Psoriasis is a chronic skin condition in which skin is shed at much higher levels than normal, with epidermal cells dividing up to 7 times faster than normal (Scott et al. 1976). Oral ingestion of 0.6 mg/kg body weight of 8-methoxypsoralen, a furanocoumarin, causes systemic photosensitivity within 2-3 hours, at which time psoriatic epidermal tissue is irradiated with near UV light (UVA) (Gasparro 1988). This treatment, known as PUVA (which stands for Psoralen-UVA) does not cure the patient of psoriasis but alleviates the symptoms for extended periods of time by inhibiting DNA synthesis (Scott et al. 1976). PUVA is the most effective treatment for psoriasis known. Furanocoumarins also have been used to elucidate the structure and function of nucleic acids (Gasparro 1988). Recently, they were found to have promise as a clinical treatment of T-cell lymphoma by exposing furanocoumarin-containing blood to UVA extracorporeally (Gasparro 1988). Furanocoumarins can be classified into two groups, angular and linear, depending upon the location of the furan ring (Brown 1979). Angular furanocoumarins are less reactive with DNA because of steric hindrance of the furan ring at the 7,8 position. Linear furocoumarins such as 8-methoxypsoralen (8-MOP), 5- methoxypsoralen (5-MOP), and psoralen affected cellular function more than did the angular compounds (Brown 1979). When linear furanocoumarins are photoactivated they form cross-linkages between DNA strands and disrupt normal cellular functions (Scott et al. 1976). Such double photoadduct reactions (cross-links) usually form between pyrimidine bases (Kagan 1993) and have affected DNA replication, RNA synthesis, and 117 DNA synthesis (Brown 1979). Disruption of normal DNA function with furocoumarins has been shown to be lethal or mutagenic to affected cells (Averbeck 1985, Brown 1979). Photoactivated DNA cross-linking with furocoumarins killed bacteria, yeast, and mammalian cells (Ashwood-Smith et al. 1981, Averbeck 1985, Brown 1979), inactivated tumor cells (Brown 1979), and induced mutations and carcinoma (Ashwood-Smith et al. 1980, Averbeck 1985, Hanawalt et al. 1981). Since double photoadducts induced large changes in DNA structure (Gasparro 1988) they were likely either to be recognized and repaired by repair enzymes (Hanawalt et al. 1981) or to be lethal to the cell (Saffran 1988). During repair of the cross-links, mutagenesis may occur (Saffran 1988). Monoadducts form if furanocoumarins bind to a pyrimidine base without cross-linking with another furanocoumarin (Averbeck 1985, Scott et al. 1976). Monoadducts were likely to avoid excision and subsequently encouraged mutagenesis during DNA replication (Hanawalt et al. 1981). Although furanocoumarins are best known for their potent photoreactivity with DNA, they also have dark-reaction capabilities and can bind with other cellular components. In the dark, furanocoumarins intercalated between DNA bases (Averbeck 1985) without forming covalent bonds (Rodighiero et al. 1988). Frameshift mutations were common in bacteria exposed to 8-methoxypsoralen doses of 10 pig/ml or higher without light exposure (Bridges and Mottershead 1977). Also, proteins were found to have high affinities for furanocoumarin binding in the dark (Midden 1988). In fact, furanocoumarin binding to cellular components in the dark or light was at least as common as binding to DNA (Midden 1988). After PUVA treatment, the majority of 8- methoxypsoralen was found in the cytoplasm and not bound to DNA (Midden 1988). 118 Furanocoumarins were capable of binding to fatty acids, proteins, and membrane receptors (Midden 1988), which may account for some of the epidermal reactions associated with photochemotherapy and photoderrnatitis. Furanocoumarin production in celery has been most commonly associated with infection by Sclerotinia sclerotiorum (Lib.) de Bary (Chaudhary et al. 1985, Scheel et al. 1963). Linear furocoumarins act as phytoalexins in celery since they are produced in response to external stresses (Beier and Oertli 1983). Increased levels of the linear furanocoumarins psoralen, 8-methoxypsoralen, and 5-methoxypsoralen have been observed in celery following infection by Fusarium oxysporum f. sp. apii race 2 (Heath- Pagliuso et al. 1992), parsnip yellow fleck virus (Ataga et al. 1993, Lord et al. 1988), and Erwinia carotovora (Karasawa et al. 1990). Exposure to acidic fogs (Dercks et al. 1990), ultra violet light, cold, sodium hypochlorite, and copper sulfate (Beier and Oertli 1983) also stimulated linear furanocoumarin production. Furanocoumarin levels in healthy celery were found to be 1.3 ppm in Florida 683 and 1 ppm in Tall Utah 52-70R (Beier et al. 1983). Fusarium oxysporum f.sp. apii infection caused up to a 17-fold increase in linear furanocoumarins over these base levels (Heath-Pagliuso et al. 1992). Linear furanocoumarins exhibited antimicrobial properties in the dark (Desjardins et al. 1989, Fowles et al. 1958) and, more commonly, following exposure to near ultraviolet (UVA) light (Ashwood-Smith et al. 1981, Stanley and Jurd 1971). It is possible that increased celery disease resistance could result if levels of furanocoumarins were elevated to fungitoxic levels. Jordan et al. (1989) found that more phenol- containing electron-opaque structures accumulated in resistant celery cultivars than in susceptible celery cultivars following FOA2 infection. These structures may provide a 119 chemical or structural barrier to Fusarium spread within the tissue (Jordan et al. 1988). Cerkauskas and Chiba (1991) found a significant positive relationship between resistance and furanocoumarin concentration. Improved pest resistance in sunflower somaclonally- derived regenerants have also been shown to be partially regulated by elevated coumarin levels (Roseland et al. 1991). Perhaps resistance to Fusarium yellows of celery in the somaclonally—derived celery lines (Chapter 2) is achieved partially through increased levels of linear furanocoumarins. It is unknown, however, whether psoralen, 8- methoxypsoralen, or 5-methoxypsoralen are toxic or inhibitory to FOA2. The objective of this research was to examine the effect of the linear furanocoumarins 8-methoxypsoralen and 5-methoxypsoralen on the growth of Fusarium oxysporum f.sp. apii race 2 in the dark and following ultraviolet irradiation. A few celery lines were also screened to determine if high furanocoumarin levels existed in the somaclones which could account for the observed disease resistance to Fusarium yellows of celery. 120 MATERIALS AND METHODS Dry weight assays Erlenmeyer flasks containing 20 ml of sterile modified Czapek-Dox solution (Tuite 1969) with only 15 g/L sucrose (half of normal) and 0.5% dirnethyl sulfoxide (v/v) with or without 8-methoxypsoralen or 5-methoxypsoralen (Sigma Chemical, St. Louis, M0) were each inoculated with 2.5 x 10° Fusarium oxysporum f. sp. apii race 2 conidia. Within three hours, half of the flasks at each dose of furanocoumarin were exposed to 366 nm ultraviolet light for one hour at a distance of 10 cm, which was the optimal exposure for production of a phototoxic response (Van der Sluis et al. 1981). Six flasks were used per dose, three of which were exposed to UVA. All flasks were wrapped in foil to exclude light and shaken at 100 RPM for 7 days. The culture suspensions were then filtered through dried and pre-weighed Whatrnan #2 filter paper and the collected fungal matter was dried at 75-80°C for 24 hours. Fungal dry weight was calculated by subtracting the weight of the filter paper from the total dry weight. Furanocoumarin doses tested were 0, 0.5, 5, 10, 50, 100, 150, and 200 jig/ml (0, 0.0023, 0.023, 0.046, 0.231, 0.463, 0.694, and 0.925 mM, respectively) for 8- methoxypsoralen and 0, 0.5, 5, 10, and 50 [Lg/I111 (0, 0.0023, 0.023, 0.046, and 0.231 mM, respectively) for 5-methoxypsoralen. Higher doses of 5-methoxypsoralen were not tested because the compounds did not completely dissolve in the dirnethyl sulfoxide. To determine if dirnethyl sulfoxide was fungitoxic to FOA2, flasks were also prepared that did not contain any dirnethyl sulfoxide or furanocoumarin. All experiments were repeated once. 1 2 1 Radial growth assays Aliquots of 0.5% dimethyl sulfoxide (v/v) in which 0, 0.5, 5, 10, 50, 100, 150, and 200 jig/ml doses (0, 0.0023, 0.023, 0.046, 0.231, 0.463, 0.694, 0.925 mM, respectively ) of 8-methoxypsoralen or 0, 1.5, 5, 10, and 50 doses (0, 0.0023, 0.023, 0.046, 0.231, respectively) of 5-methoxysporalen (Sigma Chemical, St. Louis, M0) were dissolved were each added to duplicate flasks containing 50 ml of sterile, molten (50°C) modified Czapek-Dox agar (Tuite 1969) with 15 g sucrose/L instead of the normal 30 g sucrose/L. Duplicate flasks without dirnethyl sulfoxide were also used to determine if the solvent had properties inhibitory to FOA2. The agar medium was stirred briefly and dispensed into five 1.5x9 cm Pyrex9 Petri dishes (approximately 10 ml in each). Mycelial disks (6 mm diameter) from one month old FOA2 cultures on Czapek’s agar were placed on the cooled agar with the mycelial surface appressed to the agar at the center of each plate. Within three hours, half of the plates for each dose were exposed to 366 nm ultraviolet light for 1 hour at a distance of 10 cm. For 5 days, the plates were kept in the dark at room temperature (22-24°C) at which time the colony diameters were measured. All experiments were repeated once. Statistical analysis Analyses of dry weight and radial growth were examined separately. For each experiment, 2-way AN OVA tests were completed using SigmaStat statistical software (Jandel Scientific, San Raphael, CA) to determine whether dose, ultraviolet light, or the interaction between them affected growth. Growth responses to dosages between 0 and 10 jig/ml were predicted using linear regression (P=0.05). For each study, regression 122 slopes were compared with a t-test (P=0.05) between UV and no UV treatment to determine whether the growth responses to dose were similar. Mean separation for experiments at higher doses (up to 200 jig/ml) were determined with Bonferroni’s test. Furanocoumarin extraction from celery tissue Celery plants were grown in naturally infested muck soil at Wilbrandt Farms, Decatur, MI. Five plants per line were rated for disease on a 1-5 scale, where 1 = no vascular discoloration, 2 = trace of vascular discoloration, 3 = < 50% crown discolored, 4 2 50% crown discolored, and 5 = dead or nearly dead plant. For extraction of furanocoumarins, the lines chosen were: FL 1-2, FL 2-1, FL 3-2, FL 3-3-5, K 26, F 128(3)-1, F 128(4), Tall Utah 52-70 HK, and Florida 683. Three 0.7 cm diameter X 0.5 cm thick disks were taken from the crown region of the selected celery plants. The fifteen disks per line were weighed and stored in ethanol at 4°C in the dark. The disks were later boiled in 95% ethanol for 5 minutes, homogenized at room temperature, and filtered through Whatman #2 filter paper, washing with ethanol. The filtrate was evaporated under vacuum until nearly dry and then resuspended in 1 ml ethanol. These solution were cooled to 4°C and then centrifuged at 12,400 rpm for 1.5 minutes. The liquid was removed, evaporated under a stream of N2, and resuspended in 0.05 ml ethanol per gram of original crown tissue. Ten microliters of the extracted solution (equivalent to 0.2 g celery crown tissue fresh weight) and 1 pg, 0.1 11g, and 0.01 11g of 8-MOP and 5-MOP were spotted on Fisherbrand" Redi/Plate, Silica Gel G TLC plates, developed in chloroform and observed under UV light. The colors and areas of fluorescent zones likely to be 8-MOP or 5- 123 MOP were recorded, and the plates were then bioassayed (Homans and Fuchs 1970, Van der Sluis et al. 1981) with a Cladosporium cucumerinum Ellis & Arth. spore suspension suspended in Czapek-Dox solution (T uite 1969). Plates were sprayed with the concentrated spore suspension, exposed to 366 nm ultraviolet light for 1 hour at a 10 cm distance, and incubated in the dark at high humidity and 22-24°C. Inhibition zones were measured in 2-3 days. 124 RESULTS Effect of S-MOP on dry weight of FOA2 Regression analysis did not provide evidence of any strong relationships between 5-methoxypsoralen dose and dry weight of Fusarium oxysporum f.sp. apii without an ultraviolet light exposure (Figure 7). Increasing doses of 5-MOP from 0-10 [lg/[Ill did reduce dry weight of FOA2 significantly (P=0.05) following ultraviolet irradiation. Analysis of variance indicated that ultraviolet light exposure reduced the final dry weight of the fungus significantly compared with a dark treatment in only the second study (Figure 7B). Effect of S-MOP on radial growth of FOA2 5-methoxypsoralen inhibited radial growth as doses increased with or without ultraviolet light exposure (Figure 8). Regression analysis and ANOVA testing indicated that there was a significant dose effect on radial growth with or without ultraviolet irradiation (P=0.01), but that ultraviolet light inhibited growth greater than 5-MOP without UV irradiation (P=0.05) (Figure 8). Effect of 8-MOP on dry weight of FOA2 There was evidence (P=0.05) of a dose effect on dry weight in the first study only without ultraviolet light exposure (Figure 9A) and following ultraviolet irradiation in the second study (Figure 9B). Between doses 0-10 jig/ml, there were no significant 125 Figure 7. Effect of 5-methoxypsoralen (5-MOP), with or without a one hour 366 nm ultraviolet light exposure, on dry weight of Fusarium oxysporum f. sp. apii race 2 after 7 days. Results from two studies are provided separately (A, B). 126 noUV l I I 20 30 40 S-MOP Dose (pg/ml) ‘ 50 60 (112 DJ] - (110 - (109‘- (108‘- (107-1 (106'— FOA2 Dry Weight (g) (105'— ‘— (104-— (103:4 no UV UV (102 O I l I 20 3O 40 S-MOP Dose (pg/ml) 60 127 Figure 8. Effect of 5-methoxypsoralen (5-MOP), with or without a one hour 366 nm ultraviolet light exposure, on radial growth of Fusarium oxysporum f.sp. apii race 2 after 5 days. Results from two studies are provided separately (A, B). 30 40 50 60 S-MOP Dose (pg/ml) 20 IO 0 6 O 5 0 4 \1 W m o m We n u. ( e m. m. D m P 1 O o M. 2 5 m __1._1_ 0 s1__s_____ O 5 O 5 0 S O 5 0 5 O O 5 0 5 0 5 O 5 0 5 0 7 6 6 5 5 4 4. 3 3 2 2 7 6 6 5 5 4 4 3 3 2 2 38¢ .83ng 13200 20m 995 .53ng 18200 20m 129 Figure 9. Effect of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on dry weight of Fusarium oxysporum f. sp. apii race 2 after 7 days. Results from two studies are provided separately (A, B). 130 0.12 0.11 ‘7. 0.10— . A :93 0.09 353, 0.08 g 0.07 E 0.06 2 0.05 g 0.04 0.03 '— 0.02 — 0.01 — 0-00 1 1 1 1 1 0 10 20 30 40 50 60 8-MOP Dose (pg/m1) 0.12 0.11 — 0.10 - 0.09 1 0.08 1 0.07 4 “0 UV 0.06 — 0.05 — 0.04 — 0.03 — 0.02 - 0.01 - 0-00 1 1 1 1 1 0 10 20 30 40 50 60 8-MOP Dose (rig/ml) FOA2 Dry Weight (g) UV 131 differences (P=0.05) in growth response following an ultraviolet exposure compared with no irradiation. In both studies (Figures 10A and 10B), 8-methoxypsoralen doses of 50 ug/ml and greater significantly reduced the dry weight of FOA2 compared with the dry weight produced from a zero dose treatment. Dry weights were statistically equivalent for all doses between 50 jig/ml and 200 rig/ml for each treatment with or without UV within each study (Figure 10). At 8-MOP doses of 50 rig/ml or greater, UV-exposed FOA2 yielded a significantly lower dry weight than if not irradiated for only one (Figure 10A) of the two studies. UV did not affect growth response (dry weight) in the second study (Figure 10B). Effect of 8-MOP on radial growth of FOA2 A significant (P=0.01) decrease in radial growth was noticeable at 50 rig/ml of 8-methoxypsoralen compared with all 8-MOP doses 5 10 jig/ml (Figures 11 and 12). At low doses (Figure 11) regression analysis determined that increasing doses, with or without UV irradiation, inhibited radial growth significantly (P=0.01). In the first study, UV irradiation restricted growth in the presence of 8-MOP significantly more than treatments without UV exposure (P=0.05) (Figure 11A). UV light did not contribute to radial growth inhibition in the second study at all doses tested (Figure 11B). 8-MOP doses of 50 rig/ml or greater restricted radial growth of FOA2, compared with radial growth on a medium with no 8-MOP (Figure 12). In the first study, there was more radial growth at the 50 jig/ml dose than at doses of 100 jig/ml or 150 [lg/I111 (P=0.05) (Figure 12A). Following UV exposure, no differences in radial growth among 132 Figure 11. Effect of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on radial growth of Fusarium oxysporum f.sp. apii race 2 after 5 days. Results from two studies are provided separately (A, B). 133 0.12 0H- 0.10 — . A @0w1 (gow— '20M— 0%— Qum— 00m- 0m— 0m— 0m— 0m , , o 50 100 150 200 1111:: 8-MOP Dose (11 g/ml) Dry We 0.12 0.11 '- 0.10 - 0.09 - 0.08 ~ 0.07 1 0.06 7, 0.05 - no UV .9 i 1 FOA2 Dry Weight (g) 0m— 0m— 0m— 0m , , , , 0 50 100 150 200 8-MOP Dose (pg/ml) UV 134 Figure 10. Effect of high doses of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on dry weight of Fusarium oxysporum f.sp. apii race 2 after 7 days. Results from two studies are provided separately (A, B). 3O 40 50 60 8-MOP Dose (rig/ml) 20 10 0 6 0 5 V m U W do u. ( e m m D g P 1 0 0M. 28 w 1_1_______s_0 _1_________s O memememumnmumso mememamumxmums A 0 20m 3:5 83:85 mac—co 20m 136 Figure 12. Effect of high doses of 8-methoxypsoralen (8-MOP), with or without a one hour 366 nm ultraviolet light exposure, on radial growth of Fusarium oxysporum f.sp. apii race 2 after 5 days. Results from two studies are provided separately (A, B). 137 200 150 100 8-MOP Dose (jig/ml) SO no UV UV a _ _ _ _ _ 5 w 5 O 5 0 6 5 5 4 4 AEEV ~808me €200 _ 5 3 _ _ no so 1. 9. 1 .0 n2 2 _1_ umso Om 100 150 200 8-MOP Dose (pg/ml) 50 138 100, 150, and 200 jig/ml existed. However, the 200 jig/ml dose without UV exposure produced colony diameters greater than 100 rig/ml and 150 jig/ml (P=0.05) (Figure 12A). Ultraviolet light was only found to have an effect at 200 rig/ml in the first study. In the second study, ultraviolet light had a deleterious effect on radial growth at each dose except zero (P=0.05) (Figure 12B). Within each of the treatments with or without UV irradiation, the radial growth was statistically greater in the 50 ug/ml 8-MOP dose than at higher doses. Colony diameters were statistically equivalent in the 100, 150, and 200 jig/m1 dose treatments within each UV exposure treatment (Figure 123). Furanocoumarin content in somaclones The 8-methoxypsoralen standards of 10, 5, 1, 0.5, and 0.1 ug/ml produced yellow fluorescent bands with areas (averages of 2 replicate TLC plates) of 170, 136, 54, 49, and 12 m2, respectively (Table 9). The 5-methoxypsora1en standards at 10, 5, l, 0.5, and 0.1 jig/ml were yellow-blue fluorescent bands with areas (averages of 2 replicate TLC plates) of 170, 144.5, 78, 58.5 and 12 m2, respectively (Table 9). Colors and positions on the plates of some of the fluorescent compounds from celery crown extracts were similar to the standards so that they were expected to be 8-MOP and S-MOP. The areas of the bands were fairly similar among the somaclonal lines examined (30-65 mm2), regardless of the disease ratings (Table 9). One of the isolations from the TU 52-70 HK-derived somaclones, K26 had a smaller band of 8-MOP than the other lines; the Florida 683-derived somaclone FL1-2 had the largest band of 8-MOP of all the somaclones screened (Table 9). The somaclone isolations FL1-2 and FL3-3-5A had the largest 5-MOP fluorescent bands near 60 m2 (Table 9). Severely infected 139 Florida 683 had 96 mm2 bands of 5-MOP and 65 mm2 bands of 8-MOP; Tall Utah 52- 70 HK crowns with a disease rating of 3 produced bands of 48 mm2 for 5-MOP and 44 mm2 for 8-MOP (Table 9). Standard curves were produced using the band areas of the 8-MOP and 5-MOP solutions with known concentrations. From these curves furanocoumarin concentrations of the celery crown extracts were estimated using the band areas on TLC plates. From these estimates, infected Florida 683 produced the highest levels of furanocoumarins at 14 jig/g crown tissue for 8-MOP and 21 ug/ g for 5-MOP (Table 9). Although many of the somaclonal lines appear to have elevated levels of 8-MOP and 5-MOP compared with reported levels in healthy tissue of Florida 683 and Tall Utah 52-70 HK (Cerkauskas and Chiba 1991), the amounts are not as high as those found in infected Florida 683 (Table 9). FL 1-2 had the highest estimated levels of both 8-MOP and S-MOP compared with the other somaclones (Table 9). The FL 3-3-5 line had moderately low levels of 8-MOP and 5-MOP in both extractions (A and B) despite the high average disease rating (Table 9). The K26 line expressed the lowest levels of 8-MOP and 5-MOP of all the somaclones examined. Infected Tall Utah 52-70 HK expressed quite low levels of the furanocoumarins. The bioassays with Cladosporium showed inhibition zones at the fluorescent zones thought to be furanocoumarins. The inhibition zones were similar in size to the fluorescent bands measured (not shown). Since high performance liquid chromatography was not used, the actual concentrations of the furanocoumarins in tissue are not known. 140 Table 9. Estirnations of furanocoumarin content in celery lines derived from somaclonal variation compared with parent cultivars Florida 683 and Tall Utah 52—70 HK, using thin-layer chromatography. Extract Disease 8-methoxypsoralen 5-methoxypsoralen Ratingl Zone (mm2)2 Concentration3 Zone (mm2) Concentration FL 1.2 2.0 60.6 13 pg/g 60.6 10 pg/g FL 21 1.0 40 5.5 pg/g 54 7.8 pg/g FL 3.2 1.2 40 5.5 pg/g 48 5.5 pg/g FL 3-3-5A 3.2 36 4.8 pg/g 60 10 pg/g FL 33513 3.2 45 7.5 pg/g 45 5 pg/g K26A 1.2 32 3 pg/g 32 l pg/g K268 1.2 31.5 2.8 pg/g 40 3.5 pg/g F128(3)-1 1.0 44 7.5 pg/g 54 7.8 pg/g F128(4) 1.4 36 4.8 pg/g 44 4.8 pg/g FL 683 1.0‘ - 0.7-1.8 pg/g - 1.1-2.2 pg/g '1 4.0 65 14 pg/g 96 21 pg/g TU 5270 HR 1.05 - 0.4-0.6 pg/g - 3.5 pg/g 3.0 44 7.5 pg/g 48 5.5 pg/g 8-MOP‘ - 170 10 pg - - - 136 5 pg - - - 54 1 pg - - - 49 0.5 pg - - 12 0.1 pg 5-MOP‘ - - - 170 10 pg - - - 144.5 5 pg - - - 78 1 pg - - - 58.5 0.5 pg 12 0.1 pg ‘Mean disease ratings of the 5 plants were based on a l to 5 scale, where: l = no vascular discoloration; 2 = trace of vascular discoloration in crown area; 3 = < 50 % crown area discolored; 4 = 250 % crown area discolored; 5 = dead or nearly dead plant. 2Area of the fluorescent band that was similar color and position on the TLC plates as the pure compound. 2Estimates of the concentration in the original tissue, based on areas of the fluorescent band compared with the known amounts of the compounds (0.1, 0.5, 1, 5, and 10 pg). ‘Furanocoumarin concentrations of healthy Florida 683 petioles from Cerkauskas & Chiba (1991). sFuranocoumarin concentrations of healthy Tall Utah 52-70 HK petioles from Cerkauskas & Chiba (1991). °Pure 8-methoxypsoralen and S-methoxypsoralen (Sigma Chemical, St. Louis M0) were dissolved in ethanol and run on all TLC plates to provide comparisons with plant extracts. Standards were also run twice on TLC plates to produce standard curves. 141 DISCUSSION In general, 8-methoxypsoralen and 5-methoxypsoralen restricted dry weight and radial growth more as dosage increased. A single UV exposure inhibited growth in the presence of 8-MOP or S-MOP even at low concentrations of furanocoumarins. However, the growth response following UV exposure was not always different from growth in the absence of UV for both studies in each experiment. A single ultraviolet light exposure, albeit optimal conditions for response (Van der Sluis et al. 1981), may be insufficient to cause consistent growth inhibition. Radial growth experiments were originally done by growing the plates under fluorescent lights with a 12 hour photoperiod and measuring the radial growth daily. These studies (not shown) indicated that daily low levels of UV light from fluorescent lights inhibited FOA2 growth almost completely in the presence of 50 pg/ml of 8-methoxypsoralen. Furanocoumarins were most toxic to FOA2 at high doses ( 2 50 pg/ml). However, 8-MOP doses of 50 pg/ml and greater inhibited growth to similar extents, suggesting that at doses 5 50 pg/ml the furanocoumarins may have attained maximum solubility in the media used. Regression analysis indicated that increasing doses up to 10 pg/ml of 8-MOP and 5-MOP did restrict radial growth of FOA2 with or without ultraviolet irradiation. In terms of host resistance in celery, it was important to find that furanocoumarins are toxic to FOA2 without UV irradiation. Since FOA2 is a soilborne pathogen, furanocoumarins produced in the root or crown tissue must be toxic to FOA2 without UV light exposure if they are to provide host resistance. Furanocoumarins have been 142 implicated in host resistance based on their elevated levels in resistant celery (Cerkauskas and Chiba 1991). However, the furanocoumarin concentrations required for toxicity in vitro had to be quite high to provide substantial inhibition of growth. Preliminary studies (Table 9) indicated that the furanocoumarin levels expressed in celery somaclones resistant to Fusarium yellows of celery ( s 13 pg/g) are not high enough to completely inhibit growth. If these compounds help confer disease resistance in celery, they are likely to affect growth at lower levels in vivo or help provide disease control in collaboration with other secondary compounds or defense mechanisms. Susceptible celery infected by Fusarium oxysporum f.sp. apii race 2 produced 50 pg of linear furanocoumarins per gram of fresh tissue (Heath-Pagliuso et al. 1992) which was higher than the estimates of this study (Table 9). Although these levels are high enough to inhibit growth in Vitro (Figures 7-12), they apparently do not prevent infection and disease progression in vivo. It is possible that these compounds in susceptible cultivars are not expressed early enough to prevent the growth of FOA2 into healthy tissue. In a similar system, furanocoumarins were toxic to Fusarium sporotrichioides Sherb. in vitro (Dejardins et al. 1989). This fungus caused severe disease of parsnip even though the levels of the toxic furanocoumarins in infected parsnip were well above concentrations inhibitory to the fungus (Dejardins et al. 1989). However, the concentrations of furanocoumarins in infected tissues were high only for very small areas around the lesions. The fungus was able to grow into neighboring healthy tissue ahead of furanocoumarin expression (Dejardins et al. 1989). Perhaps in celery, furanocoumarin accumulation is also too slow to prevent disease. 143 Since furanocoumarins appear to be toxic to FOA2 at high levels, it is possible that constitutive expression in celery could provide resistance to Fusarium yellows of celery. The disease resistant somaclones screened appeared to have furanocoumarin levels above that of healthy celery but the levels do not appear to be high enough to be fungitoxic. One line (FL 1-2) that was mildly infected produced furanocoumarin levels nearly as high as cultivars infected with Sclerotinia sclerotiorum (Ashwood-Smith et al. 1985). Possibly, this cultivar was moderawa resistant because of elevated furanocoumarin levels expressed or more rapid furanocoumarin expression following fungal infection. In contrast to its parent, Florida 683, the somaclone FL 3-3-5 had quite low levels of furanocoumarins despite heavy infection (Table 9). Furanocoumarin concentrations in some of the somaclones may be high enough to cause photodermatitis in the presence of ultraviolet light (Karasawa et al. 1990). This may be an occupational concern for celery harvester and grocery workers (Scheel et al. 1963, Berkley et al. 1986), although the levels are not likely high enough to warrant concern by consumers (Wagstaff 1991). LITERATURE CITED Ashwood-Smith, M.J., O. Ceska, and SK. Chaudhary. 1985. Mechanism of photosensitivity reactions to diseased celery. Brit. Med. J. 290: 1249. Ashwood-Smith, M.J., A.T. Natarajan, and GA. Poulton. 1981. Comparative photobiology of psoralens. pp. 117-131. In: Psoralens in Cosmetics and Dermatology. Pergamon Press, Paris, France. p.447. Ashwood-Smith, M.J., G.A. Poulton, M. Barker, and M. Mildenberger. 1980. 5- methoxypsoralen, an ingredient in several suntan preparations, has lethal, mutagenic and clastogenic properties. Nature 285:407-409. Ataga, A.E., H.A.S. Epton, and RR. Frost. 1993. Effect of virus infection on the concentration of furanocoumarins in celery (Apium graveolens L. var. dulce Mill. D.C.). Physiol. Mol. Plant Pathol. 42:161—168. Averbeck, D. 1985. Relationship between lesions photoinduced by mono— and bi- functional furocoumarins in DNA and genotoxic effects in diploid yeast. Mutat. Res. 151:217-233. Beier, R.C., G.W. Ivie, E.H. Oertli, and D.L. Holt. 1983. HPLC analysis of linear furocoumarins (psoralens) in healthy celery (Apium graveolens). Food Chem. Toxicol. 21:163-165. Beier, R.C. and EH. Oertli. 1983. Psoralen and other linear furocoumarins as phytoalexins in celery. Phytochem. 22:2595-2597. Berenbaum, M. 1981. Patterns of furanocoumarin distribution and insect herbivory in the umbelliferae: plant chemistry and community structure. Ecol. 62: 1254-1266. Berkley, S.F., A.W. Hightower, R.C. Beier, D.W. Fleming, C.D. Brokopp, G.W. Ivie, and CV. Broome. 1986. Dermatitis in grocery workers associated with high natural concentrations of furanocoumarins in celery. Ann. Intern. Med. 105:351-355. 144 145 Bridges, BA. and RP. Mottershead. 1977. Frameshift mutagenesis in bacteria by 8—methoxypsoralen (methoxalen) in the dark. Mutat. Res. 44:305-312. Brown, SA. 1979. Biochemistry of the coumarins. pp. 249-286. In: Biochemisg of PlfiPhenolics - Recgt, Advgces in Phytochemistry, Vol.12. eds. Swain, T., J.B. Harbome, and CF. Van Sumere. Plenum Press, New York, NY. Cerkauskas, R.F. and M. Chiba. 1991. Soil densities of Fusarium oxysporum f.sp. apii race 2 in Ontario, and the association between celery cultivar resistance adn photocarcinogenic furocoumarins. Can. J. Bot. 13:305-314. Chaudhary, S.K., O. Ceska, P.J. Warrington, and M.J. Ashwood-Smith. 1985. Increased furocoumarin content of celery during storage. J. Agric. Food Chem. 33:1153-1157. Del.eo, V.A. 1992. Photocontact dermatitis. pp. 84-99. In: Photosensitivig. Igaku-Shoin Medical Publishers Inc., New York, NY. p.181. Dercks, W., J. Trumble, and C. Winter. 1990. Impact of atmospheric pollution on linear furanocoumarin content in celery. J. Chem. Ecol. 16:443-454. Desjardins, A.E., G.F. Spencer, R.D. Plattner, and MN. Beremand. 1989. Furanocoumarin phytoalexins, trichothecene toxins, and infection of Pastinaca sativa by Fusarium sporotrichioides. Phytopathol. 79: 170-175. Fowles, W.L., D.G. Griffith, and EL. Oginsky. 1958. Photosensitization of bacteria by furocoumarins and related compounds. Nature 181:571-572. Gasparro, RP. 1988. Psoralen DNA interactions: thermodynamics and photochemistry. pp. 5-36. In: Psoralen DNA Photobiology Vol. I & II. CRC Press Inc., Boca Raton, FL. Hanawalt, P.C., J. Kaye, C.A. Smith, and M. Zolan. 1981. Cellular responses to psoralen adducts in DNA. pp. 133-142. In: Psoralens in Cosmetics and Dermatology. Pergamon Press, Paris, France. p.447. Heath-Pagliuso, S., S.A. Matlin, N. Fang, R.H. Thompson, and L. Rappaport. 1992. Stimulation of furanocoumarin accumulation in celery and celeriac tissues by Fusarium oxysporum f.sp. apii. Phytochem. 8:2683-2688. Johnson, C., D.R. Brannon, and J. Kuc. 1973. Xanthotoxin: a phytoalexin of Pastinaca sativa root. Phytochem. 12:2961-2962. 146 Jordan, C.M., L.S. Jordan, and R.M. Endo. 1989. Association of phenol-containing structures with Apium graveolens resistance to Fusarium oxysporum f.sp. apii race 2. Can. J. Bot. 67:3153-3163. Kagan, J. 1993. Organic Photochemisg: Principles and Applications. Academic Press Ltd., London. p.234. Karasawa, D., H. Shibata, N. Horiuchi, Y. Andou, and M. Simada. 1990. Photoactive furocoumarins in diseased celery (Apium graveolens). Agric. Biol. Chem. 54:2141-2142. ' Lord, K.M., H.A.S. Epton, and RR. Frost. 1988. Virus infection and furocoumarins in celery. Pl. Path. 37:385-389. Midden, W.R. 1988. Chemical mechanisms of the bioeffects of Furocourmarins. pp. 1-50. In: Psoralen DNA Photobiology Vol. II. Gasparro, F.P. ed. CRC Press Inc., Boca Raton, FL. Rodighiero, G., F. Dall’Acqua, and D. Averbeck. 1988. New psoralen and angelicin derivatives. p. 37-114. In: Psoralen DNA photobiology Vol. 1. CRC Press Inc., Boca Raton, FL. Roseland, C.R., A. Espinasse, and T.J. Grosz. 1991. Somaclonal variants of sunflower with modified coumarin expression under stress. Euphytica 54: 183-190. Saffran, W.A. 1988. Genotoxic effects of psoralen. pp. 73-87. In: Psoralen DNA Photobiology Vol. II. CRC Press Inc., Boca Raton, FL. Scheel, L.D., V.B. Perone, R.L. Larkin, and RE. Kupel. 1963. The isolation and characterization of two phototoxic furanocoumarins (psoralens) from diseased celery. Biochemistry 2:1127-1131. Scott, B.R., M.A. Pathak, and GR. Mohn. 1976. Molecular and genetic basis of furocoumarin reactions. Mut. Res. 39:29-74. Stanley, W.C. and L. Jurd. 1971. Citrus coumarins. J. Agric. Food. Chem. 9:1106-1110. Trumble, J .T., W. Dercks, C.F. Quiros, and R.C. Beier. 1990. Host plant resistance and linear furanocoumarin content of Apium accessions. J. Econ. Entom. 83:519-525. Tuite, J. 1969. Plant Pathological Methods. Burgess Publishing Co., Minneapolis, MN. p. 239. 147 Van der Sluis, W.G., J. Van Arkel, F.C. Fischer, and R.P. Iabadie. 1981. Thin- layer chromatographic assay of photoactive compounds (furocoumarins) using the fungus Penicillium expansum as a test organism. J. Chromatogr. 214:349-359. Wagstaff, DJ. 1991. Dietary exposure to furocoumarins. Regulatory Toxicol. Pharrnacol. 14:261-272.