CHANGES IN ULMUSI AMERICANA IN RESPONSE TO INFECTION BY" CERATOCYSTIS ULMI ., 3.23. 25:23}? {my Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY WILLIAM RONALD. LANDIS 1969 mm “Funk/1P: A IIIIIIIIIIQI‘IIIIIIIIIIIIII v; L ., Michi an Stat: 312 301090 9285 g 1 Univemty r3) This is to certify that the thesis entitled I CHANGES IN ULMUS AMERICANA IN RESPONSE TO INFECTION BY CERATOCYSTIS ULMI presented by WILLIAM RONALD LAND I S has been accepted towards fulfillment of the requirements for PH.D. degree in AGRICULTURE Major pro essor Date Jul I I 6 0-169 _.. _.._._--..__-. . . \ ABSTRACT CHANGES IN ULMUS AMERICANA IN RESPONSE TO INFECTION BY CERATOCYSTIS ULMI By William Ronald Landis EXperiments were designed to study the response of pathogen-free tissue of American elms, glmus americana L. infected with the Dutch elm disease pathogen, Ceratocystis Elm; (Buism.) C. Moreau. DevelOpment of disease following artificial inoculation was compared with disease deve10pment in natural infections. Oxygen uptake of pathogen-free leaf discs began to increase 5-11 days after inoculation of trees, and reached a maximum between 11 and 22 days, when oxygen uptake was as much as 80 percent higher than in controls. Thereafter, oxygen uptake decreased until it was less than controls 20-26 days after inoculation. Increased oxygen uptake was not caused by water stress since plants which had water stress had lower oxygen uptake than did normally watered controls. Conductivity of aqueous leachates from leaves was greater for inoculated than for control plants; this increased loss of electrolytes was correlated with increased oxygen uptake until maximum oxygen uptake was reached. When oxygen uptake began to decrease, conductivity began a rapid increase, reaching over 300 percent of controls when oxygen uptake was at its minimum of 30-50 percent of William Ronald Landis controls. Reduction of triphenyltetrazolium chloride demonstrated that twigs (from which the leaves were harvested for respiration and conductivity experiments) contained living parenchyma. MicrOSCOpic observation demonstrated that the twig, midvein, and larger veins in blades from leaves of inocu- lated trees were not visibly occluded during reSpiration and conductivity studies. Attempts to isolate 9. Elm; from these tissues were not successful. By using symptom deve10pment and presence of the patho- gen from leaves, twigs, and stems as criteria, colonization of naturally infected and artificially inoculated trees was studied. When trees were inoculated with approximately 400 Spores, the extent of colonization was similar to that of naturally infected trees. When large numbers of spores were used in inoculation, colonization was more rapid and extensive. This work demonstrates that physiological changes occur in pathogen-free tissue of Q. americana soon after inocula- tion with Q. Elmi. It suggests that occlusion and subsequent inability of vessels to transport water may occur relatively late in disease deve10pment, and that physiological changes may be due to a toxin in the tranSpiration stream. CHANGES IN ULMUS AMERICANA IN RESPONSE TO INFECTION BY CERATOCYSTIS ULMI by William Ronald Landis A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1969 all. In To Joanne Ho Ho ACKNOWLEDGEMENTS My sincere gratitude to all who made this thesis possible; especially Dr. John H. Hart, Dr. John L. Lockwood, Dr. Robert P. Scheffer, Dr. James W. Butcher, Miss Sharon Smith, Miss Hyla Clark, Mrs. Sandra English, and Mrs. Kayla Ager. iii INTRODUCTION . . . LITERATURE REVIEW MATERIALS AND METHODS RESULTS . . .n. . Distribution of Q. ulmi in naturally infected large trees and in inoculated small trees Oxygen uptake TABLE OF CONTENTS Location of the pathogen in test trees . Dehydrogenase activity in terminal twigs . Loss of electrolytes from leaves and twigs Tissue discoloration and vessel occlusion Isolation of toxic substances DISCUSSION AND CONCLUSIONS LITERATURE CITED iv 14 14 17 32 33 35 an 45 48 53 LIST OF TABLES Table Page 1 Recovery of Ceratocystis ulmi from leaves and twigs of Ulmus americana showing various degrees of symptom develOpment . . . . . . . . 15 2 Recovery of Ceratocystis ulmi from Ulmus americana inoculated with approximately 400 Sporespertree00000000000000.16 3 Oxygen uptake by leaf discs from Ulmus I americana inoculated with Ceratocystis ulmi or injected with water . . . . . . . . . . . . 28 4 Relation between extent of colonization of Ulmus americana by Ceratocystis ulmi and oxygen uptake of leaf discs . . . . . . . . . 34 5 Loss of electrolytes by leaves from Ulmus, americana inoculated with Ceratocystis ulmi or injeCted With Water 0 o 0 o o o o o o o o 38 Figure 10 ll 12 LIST OF FIGURES Symptomless, chlorotic, and necrotic leaves from Ulmus americanainfected with Ceratocystis ulmi Ia—f symptomless, g-i chlorotic, j necrOtiC) O O O O O 0 0 O O O O O 0 O O O 0 I 0 Oxygen uptake by leaf discs from Ulmus, americana inoculated on May 15, 1958 with ceratOClStiSUlmiooooooocoo.oooo Oxygen uptake by leaf discs from Ulmus americana inoculated on May 23, 1957 with ceratOCYStiSUlmiooococoa...ocoo Oxygen uptake by leaf discs from Ulmus americana inoculated on June 7, 1957 with CeratocyS-tis ulmi O 0 O O I O O 0 O O O 0 O O 0 Oxygen uptake by leaf discs from Ulmus americana inoculated on June 5, 1958 with Ceratocystis ulmi . . . . . . . . . . . . . . . Oxygen uptake by leaf discs from Ulmus americana inoculated on June 19, 1957 with ceratocyStisulmiooooo000000000o Oxygen uptake by leaf discs from Ulmus americana inoculated on June 27, 1957 with Ceratocystis ulmi . . . . . . . . . . . . . . . Oxygen uptake by leaf discs from Ulmus americana inoculated on July 5, 1958 with ceratOQ‘VStiS Ulmi o o o o o o o o o o 0 o o o o Conductance of leaves from Ulmusamericana inoculated on June 5, 1968 with Ceratocystis 1111111ooooooooooooooooooooo Conductance of leaves from Ulmus americana inoculated on July 5, 1958 with Ceratocystis Ulmi00.000000000000000... Oxygen uptake and conductance of leaves from Ulmus americana inoculated with Ceratocystis ulmi on June 5, 1968 . . . . . . . . . . . . . Oxygen uptake and conductance of leaves from Ulmus americana inoculated with Ceratocystis ulmi on July 5, 1968 . . . . . . . . . . . . . vi Page 12 21 22 23 24 25 26 27 4O 41 42 43 INTRODUCTION Vascular occlusions casued by gums and tyloses occur in American elm (Ulmus americana L.) infected by Ceratocystis ulmi (Buism.) C. Moreau (9, 20, 44). Several toxic sub— stances which occur in crude culture filtrates of the fungus elicit a disease syndrome said to be similar to that of Dutch elm disease (18, 22, 24, 53, 70, 72). The relative importance of vessel occlusion and toxic substances is not known in Dutch elm disease. Systemic effects in host or non-host tissues which did not contain the pathogen have been reported for other vascular diseases of plants. Veinclearing (23), increases in reSpiration (15, 55), increases in permeability (l6), and discoloration of parenchyma (54) occurred in pathogen- free tissues of tomato infected with Fusarium oxysporum f. lycopersici (Sacc.) Snyd. and Hans. Non-host scions of intergeneric (l7) and intrageneric (37, 38) grafts re" sponded to invasion of the host stock with necrosis. Photosynthesis decreased in pathogenmfree blades of cotton plants infected with Verticillium albo-atrum Reinke and Berthold (39). These data suggested that some substance was moving in advance of the pathogen, disrupting the normal metabolism of the host. The purpose of this research was to determine whether or not there are physiological changes in trees with Dutch elm disease which might be caused by a toxin. If so, at 1 what time in the disease syndrome do the physiological changes occur? Can changes be expressed in advance of the invading fungus, or do they occur only at the site of in- fection? What is the pattern of fungus distribution within the host in relation to possible changes occurring in tissues? Additional experiments were designed to compare the extent of colonization of the host under conditions of natural infection and artificial inoculation with reference to symptom expression and inoculation technique. LITERATURE REVIEW Dutch elm disease was first described in France in 1918 (28). The pathogen was carried in elm burl logs to North America (5) where it was identified in Ohio in 1930 (40). Since reaching North America, the pathogen has spread throughout most of the natural range of the principle host, the American or white elm (62). The range of this tree is demarked by the Atlantic Ocean, the Gulf of Mexico, and a line from Cape Breton Island, Nova Scotia westward through Ontario, northwest through eastern Saskatchewan, south through the Dakotas, and south-southeast through middle Kansas and eastern Texas (27). The total number of wild elms in North America when the disease was introduced was estimated at one billion. In addition to the wild trees there were approximately 500 million elms growing as ornamen- tal trees in cities (14). The smaller European elm bark beetle (Scolytus multistriatus Marsham) and the American elm bark beetle Hylurgopinus rufipes Eichoff) are the principal vectors of Q. ulmi (15). §. multistriatus first was found in North America in 1904 and has since become established throughout most of the range of the American elm (68). This insect is the most important vector of g. ulmi throughout most of the United States (46), but in Ontario, fl. rufipes is the primary vector (47). When elms larger than 7 inches diameter breast height (dbh) are growing within a few feet of each other, as in a street planting, root grafts between trees are common. The fungus Spreads from tree to tree through these grafts (67). Trees are most commonly inoculated by S. multistriatus contaminated with g. ulmi spores when the adults feed in twig crotches of healthy trees (15, 48). This process is of low efficiency with approximately 3-8 percent of the feeding wounds resulting in infection (45). Under favorable conditions, Spores which reach the xylem germinate and initiate infection (35. 44, 65). The pathogen spreads throughout the tree by means of yeast-like spores (1) or microendospores (43, 69) which are carried in the tranSpir- ation stream. The host usually dies two or three years after inoculation, but may die the year of inoculation if the pathogen moves into the main trunk and becomes systemic (12, 41). When non~infected elms are pruned late in the summer, they are attractive to S. multistriatus as brood trees. If the females are contaminated with Q, ulmi Spores, they commonly inoculate the tree when preparing their egg gallery (30). These trees usually die the year following infection (31). Many factors affect the develOpment of Dutch elm disease. Pomerleau (47), Smalley (60), and Banfield (4) reported that the disease develOps more readily in the spring and early summer. This was explained by the more rapid. movement of Spores in the large springwood vessels (2, 4, 11). The prevention or inhibition of disease deveIOpment by chemicals which inhibit the development of springwood or initiate occlusion of vessels is consistent with this view (7, 8, 10, 19. 59). The number of spores used for inocula» tion does not affect the incidence of disease, but does affect the time required for symptom develOpment (2, 58, 72). Banfield (2) reported that g. ulmi Spores are distributed to the uppermost branches of large trees rapidly when many spores were injected into the base of the trees. Pomerleau and Mehran (49) found Spores distributed into leaves within 24 hours of inoculation and Pomerleau and Pelletier (50) reported that the pathogen was easily obtained from leaves. In the latter report (50), the inoculation technique was not described, but in the former (2, 49) large numbers of Spores were used. Campana (12) and Neely (41) have criticized reports of rapid Spore movement. When trees were artificially inoculated at twig crotches, g. aim; usually colonized only the inoculated branch during the year of inoculation. After the pathogen reached the main trunk, colonization of the tree was more rapid. Experiments which demonstrated rapid colonization have been conducted by injecting large number of spores, suspended in large volumes of water, into trees. Spores are rapidly distributed to terminal branches following these injections, but the tech- nique is entirely artificial and does not measure the rate of natural colonization. Smalley (60) reported that increased nitrogen length- ened the susceptible period of young trees by lengthening the period of succulent growth. Zentmyer and Wallace (71) reported that nitrogen reduced susceptibility of large trees to disease by increasing the Vigor of the host. Schreiber (57) reported that trees 1-5 months old were immune to Dutch elm disease, but Smalley (60) was able to inocUlate young trees in the two leaf stage. He attributed the general lack of success of others to inoculating the trees too late in the season or to lack of nitrogen. Smucker (61) and Banfield et a1. (3) reported that trees which are less than 1 inch in diameter do not have recurrence of disease the season following inoculation. In trees larger than 15 inches dbh, disease recurrence was 100 percent (3). The rare recurrence in small trees was associated with growth of the fungus into the current year's growth ring (4, 61). Although g. ulmi was able to invade the new growth of stems in some cases, most disease recurrence occurred in roots where vessels of different growth rings were often contiguous (4). There are variations in susceptibility of elm clones (33, 34, 63) and in virulence of isolates of g. ulmi (24, 32, 64). The age of spore inoculum also affected disease develOpment (71). Disease deve10ped much slower when trees were inoculated in twigs than when spores were injected into the bole (ll, 12, 13). This apparently was associated with slow movement of g. ulmi proximally from the inoculation point. The mechanism by which 9, ulmi induces symptoms is not understood. Many workers have suggested that a toxin is involved in symptom deveIOpment. A water soluble substance obtained from the culture filtrate of g. ulmi caused symp- toms similar to those of Dutch elm disease on elm, tomato, maple, and snapdragon (70). Dimond (18) isolated two fractions from culture filtrates; a polysaccharide which caused upcurling of leaves, and an alcohol-soluble, ether- insoluble fraction which caused interveinal necrosis. Feldman et a1. (21, 22) and Frederick and Howard (24) found no correlation between the ability of an isolate to cause disease and its polysaccharide production in culture. Even though the polysaccharide component could cause wilting, they did not believe it to be the primary wilt inducing agent because removal of the polysaccharide did not reduce phyto- toxicity of the filtrate. They found that the highest toxin production occurred at pH 4.25 in 7 days, and that a higher pH prevented toxin formation and deactivated preformed toxin. 9, ulmi always lowered the pH of culture media, and the affect of pH on the pathogenicity of the fungus was not tested. Frederick and Howard (24) found no correlation between toxin titre and pathogenicity of different isolates of Q. Elm; upon elm seedlings. These reports (18, 21, 22, 24, 70) were based upon crude culture filtrates of Q. ulmi. Salemink et a1. (53). using more precise techniques, isolated two substances from culture filtrates which were toxic to elm cuttings. The first substance, a glyco- protein, reproduced Dutch elm disease symptoms alone or in combination with the second. The second component was not isolated in pure form; therefore, its phytotoxicity is un- known. Resistant controls were not tested. Beckman (6) found that pectin depolymerase and cellu- lases were produced, but that polygalacturonases were not produced in culture. He suggested that gums and tyloses observed in vessels could be a result of the action of these enzymes, but in 3119 experiments were not conducted. Beckman (9) later suggested that Dutch elm disease, as well as other vascular diseases, was the result of vascular dysfunction caused by gum and tylose formation behind the fungus in the transpiration stream. However, Beckman only studied resistant late summer wood. Elgersma (20) studied resistant and susceptible elm clones, and found that the resistant clones had gel formation in vessels, which caused a localization of infection soon after inoculation. Sus- ceptible plants had extensive colonization of the vascular system, followed by vessel occlusion. The first symptom of disease observed by Kerling (36) was discoloration of parenchyma. Her photographs indicated that gums occurred where many spores had lodged, and that tyloses formed where fewer spores were found. She related these events to the amount of hypothetical toxin which these Spores might produce. Gagnon (26) also found that early changes occurred in xylem parenchyma, but that these were preceeded by yellowing of pits in xylem walls. He found that gel plugs which were formed in vessels contained pectin as a main constituent, and that the plugs were associated with thickening of vessel walls (25). Wilson (69) found that isolated areas of the xylem were blocked by gums, tyloses, and hyphae, but that only a small portion of the vascular system was blocked when foliar symptoms became apparent. He stated that the fungus was found primarily in the paren- chyma tissue, and that this may be the tissue upon which resistance is based. Banfield (4) however, found 9. Elfll only rarely in parenchyma until disease symptoms were severe. MATERIALS AND METHODS Trees inoculated in 1967 were growing approximately 2.5 km southwest of Laingsburg, Michigan or on the Mighigan State University Department of Botany and Plant Pathology farm. Trees inoculated in 1968 were growing approximately 2.5 km south of East Lansing, Michigan. 9. ulmi was isolated from three different naturally infected trees in February 1967 and in April 1968. These six cultures were grown in shake culture at 2200 in a medium containing (g/l): glucose, 50; asparagine, 2; malt extract, 12; KHZPou, 1.5; MgSOu-7 H20, 0.5; FeCl 0.01. 3. Most growth under these conditions was bud cells with little hyphal growth. Inoculum was prepared by mixing bud cells from three isolates (approximately 106 cells per ml), or by diluting this mixture to a concentration of approximately 8000 cells per ml. Each of thirty American elm trees (2.5—4.0m tall) was inoculated with one drOp (approximately 400 cells) of diluted bud cell suspension of 9. Elm; on one of the follow- ing days: May 23, June 7, l7, and 27, 1967 and May 15, June 5, and July 5, 1968. Fifteen trees were water injected on each of these dates to serve as controls. The drOp was placed in a twig or branch axil 70-130 cm above ground level, where the trees were 3-6 cm in diameter. The bark was then pierced through the drOp with a No. 11 Bard-Parker lance. lO 11 When the lance was withdrawn, the drOp of spore suspension was rapidly drawn into the transpiration stream. Two in— oculated trees and one non-inoculated tree were harvested daily or every other day during the 30 days following in- oculation. Terminal leaves and twigs from these trees were used in respiration studies, the uppermost lateral leaves and adjacent twigs were used in conductivity studies, and the next lower lateral leaves and adjacent twigs were sectioned and stained. The extent of colonization of trees inoculated with 400 spores per tree was compared to that of naturally in- fected trees and to trees inouclated with 5 X 10LI Spores per tree. To study distribution of 9. Elm; in naturally infected trees, one or two branches which bore leaves showing early symptoms (slight chlorosis or necrosis) were cut from different large (>25 cm dbh) American elms at intervals throughout the spring and summer of 1967 and 1968. Leaves from these branches were divided into three symptom categories: symptomless (Figure l a-f), chlorotic (Figure l g—i), and necrotic (Figure 1 j). Leaves which contained more necrotic tissue than shown in Figure l j were included in the third category. Leaves and sections of the tWigs bearing the leaves were placed without surface sterilization on the surface of elm extract agar in petri plates. Tissues were considered free of the fungus when no synnemata characteristic of Q. ulmi were present after 14 days at 2200. 12 e 90 99 Figure 1. Symptomless, chlorotic, and necrotic leaves from Ulmus americana infected with Ceratoc stis ulmi (a—f symptomless, g-i chlorotic, j necrotic). 13 To study distribution of 9. Elm; in trees inoculated with 5 X 10LI spores per tree, each of five to ten American elm trees (2.5-4 m tall) were inoculated with one drOp (approximately 5 X 10LI cells) of a concentrated bud cell suspension on one of the following days: May 15, June 5, and July 5, 1968. These trees were harvested after some of their leaves began to show yellowing or browning and colonization was studied by the same techniques used for naturally infected trees. RESULTS Distribution of C. ulmi in naturally infected large trees and in_inoculated small trees - The rate and extent of colonization of Q. americana by 9. Elm; varied with different inoculation techniques (2, 12, 41, 49, 50). When many spores (>106) were used, colonization was more rapid and extensive than with natural colonization (2, 49, 50). Therefore, an attempt was made to compare colonization of trees inoculated with few spores with colonization of naturally infected trees. The number of attempted isolations from leaves and twigs with each type of symptom and the percentage of attempted isolations which contained 9. ulmi are shown in Table 1. Data are presented for naturally infected large trees, and for small trees inoculated with 5 X 10LI spores per tree. A comparison of these data with those obtained from trees inoculated with 400 Spores per tree (Table 2) indicates that the extent of colonization of trees inoculated with 5 X 10” spores per tree was more rapid and extensive than the colonization of naturally infected trees or trees inoculated with 400 spores per tree. When trees were naturally infected or when trees were inoculated with 400 Spores per tree, symptomless leaf blades rarely contained 9. glmi (0.6 and 0.0 percent respectively). Eight percent of symptomless leaves from trees inoculated with 5 X 10LI spores per tree contained 14 15 Table 1. Recovery of Ceratocystis ulmi from leaves and twigs of Ulmus americana showing various degrees of symptom development. Naturally Infected Inoculated with 53X lOLI Spores/Tree Leaf Symptoms Blade Petiole Assoc.a Blade Petiole Assoc.a Twig Twig Symptomless No. Positiveb 2 16 54 61 87 170 No. Tested 35 356 167 759 759 330 % Positive 0.6 4.5 32.3 8.0 11.5 51.5 Chlorotic b No. Positive 1 9 27 16 30 53 No. Tested- 111 111 76 140 169 81 % Positive 0.9 8.1 35.5 11.4 17.8 65.4 Necrotic b No. Positive 9 29 96 69 107 182 No. Tested 184 184 186 390 390 251 % Positive 4.9 15.8 51.6 17.7 27.4 72.5 a A portion of the twig within 2 cm of the node to which the test leaf was attached. b Number of sections from which 9. ulmi was isolated. 16 Table 2. Recovery of Ceratocystis ulmi from Ulmus americana inoculated with approximately 400 spores per treea. Area isolated Surface Sterilized Not Surface Sterilized (cm above Numberb TotalC %5 Numberb Totalc %d point of inoc.) Positive Positive 0'30 42 97 43.3 23 24 95.8 30-60 34 97 35.1 16 24 66.7 60-90 28 97 28.9 14 24 58.3 90-120 22 97 22.7 10 24 41.7 Twig 8 97 8.2 4 24 16.7 Petiole 4 194 2.1 0 48 0.0 Blade 0 194 0.0 0 48 0.0 a Isolations were attempted from the 121 trees used in oxygen uptake and conductivity studies. All leaves from which isolations were attempted were symptomless. Number of sections from which 9. ulmi was isolated. C Number of sections which were sampled for the presence of g. ulmi. d Percentage of sections from which 9. ulmi was isolated. 17 the pathogen. A similar situation occurred with chlorotic leaf blades where 0.9 percent of the lamina from natrually infected trees contained the fungus whereas 11.4 percent of the chlorotic leaf blades from trees inoculated with 5 X 10LI spores per tree contained the fungus. Colonization of trees inoculated with 400 Spores more closely resembled natural colonization than did colonization of trees inoculated with 5 X 10LI spores. Therefore, respira- tion and conductivity studies were conducted on trees which were inoculated with 400 spores per tree. Oxygen uptake - If physiological changes occurred in host tissue before marcos00pic symptoms deve10ped, they could be interpreted as being caused by a toxin moving in the trans- piration stream in advance of the pathogen. Such changes might be reflected in oxygen uptake by diseased tissue. Therefore, an attempt was made to determine whether or not oxygen uptake changed during development of Dutch elm disease. Oxygen uptake was determined with a Warburg apparatus using standard manometric technique (66). Four to six terminal leaves from each tree were selected for uniformity in size and appearance. Eighty to 100 leaf discs 7 mm in diameter or 80 to 100 twig sections 5 mm long were harvested from the terminal leaves and twigs of each tree. The discs and sections from each tree were placed on moistened filter paper in separate petri dishes until they were put into manometric flasks (approximately 1 hr after trees were cut). Twenty leaf 18 discs or stem sections from each petri dish were randomly selected and placed in each of three manometric flasks on a filter paper circle moistened with 0.2 ml distilled water. Center wells contained KOH (20 percent) to absorb evolved 002. After a 20 min equilibration period at 250C, oxygen uptake was recorded for 1-2 hr with readings taken at 15 or 20 min intervals. Experiments were run in either reduced light or in darkness to prevent interference from photo- synthesis. Leaf discs and twig sections were removed from the flasks, dried at 95°C for 48 hr, and then weighed. Data were expressed as,nl oxygen uptake per mg dry weight of tissue per hr, and as a percentage of the oxygen uptake by the non-inoculated control tree used on each day. A least squares regression analysis was performed on oxygen uptake increases using the Michigan State University Agricultural Experiment Station STAT Series Description Number 7. Input values were obtained by dividing the amount of oxygen (ml per mg) taken up in each flask by the tissue from inoculated trees by the mean amount of oxygen taken up by tissues from non-inoculated trees harvested on the same day. An analysis of variance was performed between trees harvested each day using the Michigan State University Agricultural Experiment Station STAT Series Description Number 13. Input values consisted of the amount (pl per mg) of oxygen taken up by the tissue contained in each flask. Oxygen uptake by leaf discs from inoculated trees was 19 similar to that of control trees for the first five to eleven days after inoculation (Figures 2-8, Table 3). At this time oxygen uptake by the discs from inoculated plants began to increase until it reached a maximum 11-22 days after in- oculation. Maximum oxygen uptake by leaf discs from inoc- ulated trees was as much as 80 percent higher than uptake by healthy control tissues. Oxygen uptake then began to decrease, until it was as little as 30-50 percent of control oxygen up- take when the experiments were terminated. Oxygen uptake of healthy tissue was much higher (3-5‘p1 per mg dry weight per hr) soon after bud break than when the leaves were fully expanded (0.8-1.5‘pl) (Table 3). Oxygen consumption was highly variable between trees and between days in late May and early June (Figures 2-4, Table 3) in both inoculated and control trees, but showed much less variability in later experiments (Figures 5-8, Table 3). The significance level between inoculated and control plants was <0.05 when there was more than a 20 percent change in oxygen consumption. Least squares analysis demonstrated that change in oxygen uptake curves followed the general equations y=a+bx+cx2+dx3 of y=a+bx+cx2 where x is days after inoculation (Figures 2-8).. In all cases but one, the sig- nificance probability of the curve was <0.0005 and r values were >0.72. The experiment begun on June 7, 1967 (Figure 4) had a low (0.52) r value and a significance probability of 20 0.017. The poor fit and relatively high significance level of this experiment are probably due to the high oxygen uptake (2.5 and 2.9 pl per mg dry weight per hr) of the non-inocu- lated trees on days 6 and 7. This was a twofold increase over non-inoculated trees harvested on days 5 and 8. The factor responsible for this great increase in oxygen uptake is unknown. In 1967, a seasonal effect on oxygen uptake occurred. Plants inoculated on May 23 had a significant (0.05 level) increase in oxygen uptake 7 days after inoculation (Figure 3), those inoculated on June 7 had a significant increase after 9 days (Figure 4), those inoculated on June 19 had a significant increase after 18 days (Figure 6), and those inoculated on June 27 had a significant increase after 16 days (Figure 7). This is consistent with reports (47, 60) that Dutch elm disease develops more rapidly in spring and early summer.- In 1968, oxygen uptake increase did not occur as quickly after inoculation in May as after inoculation in June and July. Dutch elm disease development has been shown previously to vary with soil moisture (35, 47, 72). Lack of a seasonal affect in 1968 could have been due to the amount of rain received in the Spring and early summer of 1968. While 8.97 inches fell during May-July, 1967, 15.13 inches fell during May-July, 1968. Oxygen uptake of twigs was highly variable between trees and in no case did a discernible pattern develOp. 21 .AHm>oH mo.ov Pascawficwwm ma Homecoo map scum mononowwac psooaom zpcoss .mm.oun .mxdoo.o+xm:o.on:mm.oum .669» you mopdoaagmu mouse spas moonp cap mo some on» mvcmmouamu pcfiom scam .wEH: mfipmhoovmuoo spas mwaa .ma an: no powwHSoocH madcaposm mafia: Scum momfic wmoa an mxmpgs :wmhxo .m madman cote—30:. 3:0 930 on mm on m p or m 09 OSI OOI OOZ (louuoa ,o z) eqmdn uaBAxo OSZ .AHo>oH mo.ov VSHOfiMHc 22 swam ma Houpcoo map 809% mocosommfic Psoonon 592638 .mm.ouu .mxmoo.o -mxmmo.o+xuom.oummm.aus .omhp you mopdoaagou oops» new; moonp 039 90 came one mpComougon pcfiom 20mm .fieas mfipmwookuoo spa: soda .mm he: so covwazoocfi undefiuoew mafia: Scam nomad mama an oxwpms cowzxo auto... 90.: .2 .0 «>00 on mm on m— or m .m onsmwm OOI- OS OSI "canon ’02) eunuch ueBAxo 093 003 23 . Ho>mH mo.ov PCmOfimficmfim ma HopPCOO map 809% mocmpmmmwp pudenda hPsze .Mm.oum .mxzoo.oumxamo.o+xmdm.o::ww.au> .omup pom mmpwofiammh moms» Spa; moose 039 mo cams map mesommmmon pawom comm .HEHS wflpmhoopmhmo spas mwma .5 6:56 Go voPmHSoocfl madcapmem mafia: Scum momfip mood as oxmpms :mmhxo .3 mssmfim cot—.3905 .030 gun on mm on mp or m o / / \ r OSI cm 09 (louuoa ’0 z) aqmdn uefixxo 003 093 24 .AHm>mH mo.ov pancakficwfim ma Homecoo 62? Scam mocmpmmmfic Pamoumm zpcmse .ww.oup .mxzoo.ouxdoa.o+owm.ouh .mmnp pom mopdoaammp mouse spa; mommy 03¢ mo came one mesommnmon Pcfiom comm .AEHS mfipmaoopmhoo spas mood .m mama so copmHSoocH demeanoem mafia: scum mowfiu mama an mxmpms nowhxo .m mpswfim :0=0_:u0:_ 3:0 m>0o on mm on m. o. m OS (’ouuoa ,0 7,) 0104‘!“ “95"‘0 OOZ OSI OOl OSZ 25 .AHo>mH mo.ov endowmacwwm ma Homesco one scum mocoumwmfic Psoopon mesoae .mu.oun .waao.o+xmma.onam:.auz .oouv pom mmpmowagos mmpnp sew; moose 03p mo some one upcomopmms Pcfiom scam .fleH: mfivmzoopmhmo new; soda .ma 6236 :o omva500cfi manoeuoEm mafia: Scum momfic mama mp mxmpms cmmzxo .w onsmflm 2020.30,: .050 «>09 on mm on mp or m 09 OSI OOI OOZ (IOJIUOD 10 z) aqmdn ueBAx'o OSZ 26 .AHm>oH mo.ov pawoamficwfim m“ Houvcoo map Eonm monoanMfiu ononmm apnoea .wm.oup .mxmoo.onmxooo.o+xmmw.onomm.muz .ompp non mopmowaaou mouse spas momma 03» Mo came on» mesmmonnmp wagon seam .«EHS mwmmwmmmmmmu new; nwma .mm case so cmpmHSOOCfi mcwoahoem mafia: scam momfic mama an oxwpms cowhxo .m musmfim :0:0.30:_ 50:0 «>0n on ma ON mp or m 09 OOl / \ \ 09!- 008 “ouuoa ’0 7;) onusdn uaBAxo 09?. 27 .AHm>oH mo.ov 9:80fimficwfiw ma Hoppzoo map 809% mononmmmwc pcmonog >pCmBE .Hw.ouu .Nxmoo.ouxmma.o+fimm.ou> .069» “mm mmpmoaagou mouse spa; moose 039 mo came one mpcmmohmop pcflom scam .fiEHs mwpmhoOPwnoo spas mwma .m maze so UumHSOOCw undefinoem used: 8099 woman mama >9 mxmpms Qmw>xo .m onswfim :0: 0_ 30.: 50:0 «>00 on mm ON 9 o— m 8?...) chalua OSL 00 “canon 10 z) exp; OOZ OSZ 28 Table 3. Oxygen uptake by leaf discs from Ulmus americana inoculated with Ceratocystis ulmi or injected with water. Date Days after Oxygen uptakea % of control Inoculated Inoculation inoculatedb controlc May 23, 1967 2 2.7 3.2 84.4 3.2 100.0 3 3.4 3.2 106.3 4 3.6 5.2 69.2 3.5 57.3 5 4. 3.2 137.5 3.3 103.1 6 3.1 3.7 83.8 3.5 3.7 94.6 7 .5 3.2 140.6 4.4 137.5 8 4.3 2.9 148.3 4.1 141.4 9 3.8 2.5 152.0 11 4.1 2.0 205.0 4.3 215.0 12 3.3 2.0 165.0 3.0 150.0 14. 3.3 1.5 220.0 1.8 120.0 June 7, 1967 2 1.6 1.6 100.0 1.7 106.0 4 2.5 2.2 113.6 1.9 86.4 5 1.5 1.2 125.0 6 1.7 2.5 68.0 1.8 72.0 7 2.3 2.9 79.3 2.5 86.2 8 1.4 1.4 100.0 1.5 107.1 9 3.3 1.5 220.0 2.3 153.3 12 1.7 1.2 141.7 1.7 141.7 13 1.0 1.2 83.3 1.1 91.7 14 1.2 1.2 100.0 1.3 108.3 .nlne 19, 1967 3 1.0 0.9 111.1 1.1 122.2 4 0.9 0.9 100.0 0.8 88.9 29 Table 3 continued Date Days after Oxygen uptakea % of control Inoculated Inoculation inoculatedb controlc 7 1.0 1.0 100.0 1.1 110.0 lLI’ 007 009 7708 1.1 122.2 16 0.8 0.7 114.3 0.8 114.3 18 1.4 0.9 155.6 1.3 144.4 0.9 150.0 25 0.7 0.6 116.7 0.7 116.7 June 27, 1967 7 0.7 0.8 87.5 0.9 112.5 9 0.8 0.8 100.0 1.2 150.0 13 0.8 0.9 88.9 1.0 111.1 15 1.0 0.8 125.0 0.9 112.5 16 1.5 0.9 166.7 1.4 155.6 20 1.1 0.9 122.2 1.1 122.2 22 0.8 0.9 88.9 1.0 111.1 May 15, 1968 5 5.5 4.8 114.6 308 7902 7 5.2 5.1 102.0 4.0 78.4 9 3.7 3.7 74.0 2.9 58.0 11 3.5 3.7 94.6 2.6 70.3 13 2.7 2.6 103.8 3.3 126.9 15 3.2 3.3 97.0 3.9 118.2 17 3.2 3.5 91.4 3.3 94.3 19 3.5 2.3 152.2 2.8 121.7 21 2.2 1.5 146.7 1.7 113.3 23 4.2 2.2 190.9 3.2 145.5 June 5, 1968 5 1.8 1.7 105.9 1.8 105.9 30 Table 3 continued. Date Days after Oxygen uptakea % of control Inoculated Inoculation inoculatedb controlC 7 2.2 1.9 115.8 2.4 126.3 9 2.2 1.6 137.5 2.6 162.5 11 1.5 1.4 107.1 1.8 128.6 13 1.7 1.3 130.8 1.7 130.8 15 1.9 1.4 135.7 1.8 128.6 17 1.6 1.2 133.3 1.8 150.0 20 1.6 1.6 100.0 2.1 131.3 22 1.6 1.4 114.3 1.6 114.3 25 1.5 1.7 94.1 1.8 105.8 26 1.2 1.4 85.7 1.2 85.7 28 0.7 1.8 38.9 0.6 33.3 July 5, 1968 3 1.7 1.7 100.0 1.6 94.1 5 1.3 1.2 108.3 1.2 100.0 7 1.1 1.3 84.6 1.2 92.3 9 1.2 1.1 109.1 1.2 109.1 11 1.1 0.9 122.2 1.1 122.2 13 1.4 1.2 116.7 1.8 150.0 17 1.3 1.0 130.0 1.3 130.0 20 1.3 0.7 185.7 1.0 142.9 21 1.5 1.3 115.4 1.8 138.5 24 0.8 1.2 66.7 1.0 83.3 26 0.8 2.1 38.1 1.2 57.1 30 0.4 1.3 30.8 aft 0.3 23.1 )L} per mg dry weight per hr. There was 50-70 mg leaf b tlssue per flask. Mean of three samples per tree, trees inoculated with 400 Spores. C €38xn of three samples per tree, trees injected with water. 31 Because changes in oxygen uptake might be due to lack of water in leaves and twigs (42, 56), two experiments were conducted on droughted trees grown in the greenhouse. American elms l m tall were transplanted from the field to 25 cm clay pots and held in a non-airconditioned greenhouse without supplementary light. The soil in which the greenhouse plants were growing was brought to field capacity by daily watering unless otherwise noted. In the first experiment, test plants were not watered after oxygen uptake measurements were begun. In the second, plants were watered daily with approximately one-half (200 ml) the water given to controls, beginning 10 days before oxygen uptake measurements were started. This had no visible effect upon the plants. They were given an additional 200 m1 7 days after measurements were begun. Oxygen uptake was determined every other day on randomly collected leaves from four test and four control plants by standard manometric technique (66). Plants with water stress did not have increased oxygen uptake compared to non-water-stressed plants. In both experiments, by the third day after watering was stOpped, oxygen uptake of leaf tissue had decreased to approximately 80 percent of that of leaf tissue of normally watered plants. Oxygen uptake continued to decrease until 14 days after watering was stOpped, when it was approximately 50 percent of that of normally watered plants. After 16 days, some of the leaves had fallen off of the water-stressed plants, and the 32 experiments were terminated. Results of oxygen uptake experiments indicate that pathogen-free leaves from trees inoculated with Q. ulmi re- spond to infection by increases in oxygen uptake. Leaves from inoculated plants did not show visible symptoms until after oxygen uptake had reached a maximum. Oxygen uptake in— creases were not due to water stress within the leaves. Therefore, changes in oxygen uptake must be an early response to infection by 9. Elm; and must be due to some factor other than water stress. Seasonal variation in oxygen uptake cor- responds to seasonal develOpment of disease and suggests that rate of oxygen uptake is closely related to disease deve10p- ment. Location of the pathogen in test trees - To determine the location of g. ElEl in trees used in respiration and con- ductivity studies, isolations were made at various locations throughout the trees. Isolations were made 0-30, 30—60, 60- 90, and 90-120 cm above the point of inoculation, from terminal twig sections, and from two leaves below those used in the respiration and conductivity studies. Tissues explants were placed directly on elm extract agar (29) in petri plates or they were dipped in alcohol, flamed, and placed on elm extract agar. Plates were observed after 10 days, and if negative, a second time after 14 days. Samples were considered negative unless synnemata characteristic of 9. Elm; were present after 14 days. 33 In all experiments conducted in 1967 and in the first experiment in 1968, all tissues were flame sterilized before they were placed on elm extract agar. Percentage of recovery of 9. Elm; from these tissues was lower than expected (Table 2). Therefore, throughout the second and third experiments of 1968, isolations from one of the two inoculated trees were attempted without prior sterilization. Isolations were attempted from the second inoculated tree with prior steri- lization. g. Elfli was recovered from approximately twice as many samples when the tissues were not sterilized (Table 2). With the exception of four petioles, g. ulmi was not recovered from leaf tissue. A general pattern of colonization relative to oxygen uptake changes was found (Table 4). The fungus was restricted to the inoculated area until oxygen uptake increased over control oxygen uptake. The maximum distance above the point of inoculation from which the pathogen was isolated varied between 60 and 120 cm after oxygen uptake increases began but before maximum oxygen uptake was reached. After maximum oxygen uptake increase, the pathogen was found in terminal twigs. Attempted isolations demonstrated that Q. ulmi was not present in leaf discs used in respiration studies. Coloni- zation of trees inoculated with 400 spores was much slower than has been previously reported for trees inoculated with larger numbers of spores (2, 49). Dehydrogenase activity in terminal twigs - If a toxin was 34 Table 4. Relation between extent of colonization of Ulmus americana by Ceratocystis ulmi and oxygen uptake of leaf discs. Maximum height above inoculation point Oxygen uptake Days after from which 9. ulmi (relative to control) Inoculation was isolated Equal to control 0 to 5-11 30 cm IncreaSing 5-11 to 11-22 60-120 cm Maximum 11-22 120 cm-terminal Decreasing from maximum Less than control 11-22 to 20-26 20—26 to 20 twigs Terminal twigs Rarely into petioles 35 present in the transpiration stream, cells in terminal twigs might be killed in advance of the pathogen. Dehydrogenase activity, a characteristic of living cells, was used as a criterion to determine whether cells were living or dead. Bark was peeled from two-three terminal twigs from each tree used for respiration and conductivity studies. Four or five pieces 1 cm long were removed from the debarked twigs and put into stoppered test tubes containing 5 m1 aqueous triphenyltetrazolium chloride (TTC) (1 percent). Test tubes were placed in a 250C incubation chamber without light. After 24 hours the piecesxmae removed and hand sections were prepared for microscopic observation. Sections were placed in water under a coverglass and observed at 100 and 430 X magnification. Cells were considered to be alive if they turned red in 24 hours, indicating reduction of the TTC to the formazan form. Twigs killed by autoclaving served as checks. Throughout the respiration and conductivity studies, cells in terminal twigs had dehydrogenase activity. The red color of formazan was localized in xylem parenchyma tissue. Apparently cells in terminal twigs were not killed during the conductivity and oxygen uptake experiments. How- ever, no quantitative estimates of activity were made and activity could have been reduced. Autoclaved twigs did not reduce TTC. Loss of electrolytes from leaves and twigs - If host 36 cell walls or membranes were damaged in the development of disease, a change in loss of electrolytes might occur in host tissue. If conductivity of aqueous leachates increased before macroscopic symptoms deve10ped, the effect might be attributed to a toxin moving in the transpiration stream in advance of the pathogen. Therefore, conductivity of aqueous leachates from pathogen-free tissues was measured during develOpment of disease. Loss of electrolytes was determined by measuring the conductivity of aqueous leachates from leaves and twigs. An Industrial Instruments Inc. Model RC-16Bl conductivity bridge with a conductivity cell having a cell constant of 1.0 or 0.1 was used. Leaves adjacent to those used in respiration studies were selected for uniformity in size and appearance, washed twice in distilled water, and three times in deionized glass distilled water. Six leaves were harvested per tree. Two of these leaves were put into each of three 120 ml beakers which contained 40 ml of deionized glass distilled water or 40 m1 of an aqueous solution of chloramphenicol. Chloramphenicol (3 ppm) did not increase the conductivity of the bathing solution. The leaves were placed in the leaching fluid as soon as the last leaves were washed (approximately 2 hr after trees were cut). Twigs were prepared in a similar manner, with five pieces 3 cm long per beaker. Beakers were maintained at 2200 on a reciprocal shaker (6.5 cm stroke, 60 cycles per min) to insure wetting; 37 after 8 hours, readings were taken. Leaf and twig sections were then dried 40 hr at 9500 for determination of dry weights. Conductivity was measured after 8 hr and expressed as micromhos per 100 mg dry weight of tissue and as a percentage of loss of electrolytes from healthy trees. A least squares regression analysis was performed on conductance increases using the Michigan State University Agricultural Experiment Station STAT Series Description Number 7. Input values were obtained by dividing the conductance (mhos per mg) of the aqueous leachate in each beaker containing tissue from inoculated trees by the mean conductance of leachates containing tissue from non-inoculated trees harvested on the same day. An analysis of variance was performed be- tween trees harvested each day using the Michigan State University Agricultural Experiment Station STAT Series Des- cription Number 13. Input values consisted of conductance measured for each beaker in mhos per mg. Conductivity of leachates from leaves was highly variable between trees and between days when water alone was used as the leaching fluid. This may have been due to multiplication of bacteria within or upon host tissue during the 8 hr test period. When chloramphenicol was added to the leaching fluid, variability between trees was reduced, but remained great (Table 5). A difference of 30 percent between values for inoculated and uninoculated plants was required for statistical significance (0.05 level). The results of the 38 Table 5. Conductance of aqueous leachates of leaves from Ulmus americana inoculated with Ceratocystis ulmi or in- jected with water. Date Days after Conductancea % of control inoculated inoculation inoculatedb controlC June 5, 1968 5 2.15 3.37 63.8 3.96 117.5 7 2.31 3.82 60.5 3.36 88.0 9 5.28 4.78 110.5 6.32 132.2 13 2.91 2.25 129.3 4.02 178.7 15 4.31 3.35 128.7 5.01 149.6 17 2.57 2.33 110.8 3.44 147.6 20 5.22 3.00 174.0 4.20 140.0 25 6.44 1.80 350.0 4.60 250.0 26 21.15 3.58 590.8 4017 11605 28 3.56 1.26 282.1 4.29 339-7 July 5, 1968 3 4.23 2.97 142.4 2.66 89.6 5 6.33 4.38 144.5 4.90 111.9 7 4.73 4.14 114.3 4.78 115.5 5.01 149.6 11 2.78 2.53 109.9 5.10 201.6 13 9.10 6.15 149.3 6.32 102.8 17 5.50 5.09 108.1 5.23 102.8 20 4.04 2.32 174.1 3.11 134.0 21 4.28 3.70 115.6 5.28 142.7 24 4.80 2.47 194.3 10.97 444.1 26 2.60 148 175.7 6'35 42901. ”fir a Mhos X 10‘"6 per g dry weight per 8 hr. p 0 Mean of three samples per tree, two trees each day. Trees were inoculated with 400 spores. Mean of three samples per tree, one tree each day. Trees were injected with water. 39 two experiments in which chloramphenicol was used were very similar to each other (Figures 9, 10). Electrolyte loss of leaves from inoculated plants closely followed that of control plants for eight days after inoculation, then it increased to about 150 percent of that of the controls in 11 or 13 days. The increase was followed by a decrease to 129 percent or 105 percent 17 days after inoculation. Electrolyte loss then increased until it was greater than 300 percent of that of controls 24 or 26 days after inoculation. Least squares analysis demonstrated that changes in electrolyte loss followed the general equation 3 y=a+bx+cx2+dx where x is days after inoculation. In both cases, the significance probability of the curve was <0.0005. No seasonal effect on loss of electrolytes was noted. No statistically significant changes in conductivity of aqueous leachates from twigs of inoculated plants could be detected due to great variability between individual trees. Results of conductivity experiments indicate that leaves from trees inoculated with 9. gig; reSpond to infection by increased loss of electrolytes. Leaves from inoculated plants did not show visible symptoms until the conductivity of aqueous leachates from inoculated plants was at least 50 per- cent greater than that from healthy plants. A rapid increase in conductivity of aqueous leachates from leaves simultaneous with a decrease in oxygen uptake by leaf discs (Figures 11, 12) occurred as chlorosis and browning of leaves deve10ped. 40 .AHo>mH mo.ov Pcwowmwcwfim ma Houpcoo one Eonm 6006060000 pudenda apnflze .H6.oup .mxaoo.o+~xomo.o-smw.o+mms.Hus .mmup 066 mmpmoaaamu means new; moons cap mo some one upcomonnou Pawom scam .flEHs mwpmaOOPmumo new; wmma .m 603% so oopda:oocw daddduuad.udadm Scam wo>dma 00 60069000000 .0 on: Hm :0_.0_:u0:_ 3:0 «>00 on ma ow mp o. m 09 OOZ OSI OOl “0qu: ’0 °/,) eauopnpuog OSZ 41 I .Aao>oa mo.ov Padoamwcwwm ma Hopvcoo on» scam monouomwfip psoosmm spuaaa .mm.ouu .mxaoo.o+~xmmo.o-xmem.o+e:m.ous .omep ewe mmpmoaagmn 669:0 :90; moon» 0:» mo came on» menomoumop psaom comm .fiEH: mwpmaoopwuoo spas m ma .m hand so coPmHSOch unwoauosm msaflm Scum mm>woa mo mocvaSpCOO .OH on: am 0020.305 330 «>00 on ma ON mp o— m 05 OOZ OSl OOL “canon 50 z) eaunpnpuog OSZ 42 undefinosm mesa: Eopm mo>woa mo mocmposccoo can exceed smw>xo c3033.: 3:0 «>00 .mmma .m wash :0 “EH3 mHPmHoopmnmo new; umpmHSoocfi .HH museum OS 0m mu ON 9. 0.. m 3 .0 u P .. n a co. m u \ a. IAN. II«1J1II 0 \...... ”/4 .......ul.|.....“l .. .. .01 \\ .....Il..l..h.\\\\ a \\ .u. x .1. \ 0 \ Id \ \ \ \ \ \ \ \ OOI. OSl OOZ (louuoa 50 z) enmdn ueBAxo OSZ 43 .mwma .m hash so Haas mfipmaoopwumo spas copwHSQOCH wcwofiaoew wsewm Scam mm>mma mo mocwpospcoo can 6x690: cmw>xo .NH mpswam 20:0. 30:. .. 2.0 «>00 OM mm ON mp Ow m 3 0 A0 If X m. m m 000000 Md 8 m u u 00000. IP M a 0000 000‘?! I“ O ‘ ) 000 II I I I 09000 \ O on w \ ........ I I I I «5...... \ \. .\ \ 9 0000 ‘0 I III \ ) i \ con-Illa 0000.0 l ’0 a \ IIIII g 0 0V nU.O “u \ I, ... c a 0 x Z w ( c m .. x m \ .HU \ R \ nu \ x 44 This phase of the disease was probably due to a general disorganization of leaf cells and is additional evidence for physiological changes preceding the presence of the pathogen. Tissue discoloration and vessel occlusion - If Dutch elm disease symptoms were caused by a toxin, damage to cells and tissue discoloration might occur in terminal leaves and twigs before macros00pic symptoms appeared. If symptoms were caused by vascular occlusion, such as gums and tyloses, these might be detected by micros00pic observation. There- fore, sections from leaves and twigs proximal to those used in respiration and conductivity studies were observed under a microscope. Possible discoloration of terminal leaves and twigs and visible vessel occlusion were studied by clearing sectioned leaves and twigs using a modification of Scheffer and Walker's method (54). Leaves and twigs from all trees used in respira- tion and conductivity studies conducted in 1967 were cut into 2-3 cm long pieces with a razor blade, vacuum infiltrated with aqueous basic fuchsin for 5 min, then bathed in aqueous basic fuchsin for 24 hr to stain the xylem. The sections were cleared in a sequence consisting of hot 85 percent ethanol for 4-5 min, 95 percent ethanol for 15-20 min, anhydrous ethanol for 30 min, and three changes of lacto-phenol over a 72 hr period. The sections were then preserved in xylol until further sectioning. Ten to 20 transverse and longitudinal sections were prepared from each tree with a sliding microtome 45 at 20-120 p, or by hand with a razor blade. Sections were mounted in diaphane under a coverglass and observed at 50-970 X magnification. Leaves from inoculated trees deve10ped macros00pic symptoms approximately 20 days after inoculation, when they became chlorotic. Parenchyma and vessels of twigs and leaves from inoculated trees could not be distinguished from those of control trees during the 30 days after inoculation. Vessels of twigs, petioles, and midveins were free from visible occlusions with one exception. One petiole from an inoculated tree was found in which all vessels were blocked by a brown granular mass. In no case wasiflmafungus detected by micro- scopic observation. Vessel occlusions and the fungus might have been removed by sectioning and staining technique. The possibility of mechanically removing occlusions was reduced, however, by preparing thick sections. These observations indicated that visible vascular occlusions in the small twigs and leaf petioles were not responsible for symptom develOpment. Isolation of toxic substances - Respiration, conductivity and colonization studies suggested that a toxic substance might be causing physiological changes in pathogen-free tissue. Therefore, an attempt was made to isolate a toxic substance from culture filtrates of 9. Elm}. A modification of the method of Pringle and Braun (51) was selected. Three isolates of Q. ulmi used for respiration 46 and conductivity studies were grown separately in still or shake culture at 220C for 10 or 20 days in 250 ml flasks, each containing 50 ml of the medium used in preparation of inoculum or 50 ml of yeast extract supplemented (1.0 g) Fries No. 3 basal medium (51). The shaker was set for a 6.5 cm stroke and 60 cycles per min. In one experiment, medium from 10 day cultures grown in still culture in modified Fries medium was tested. In a second, media from 20 day cultures of the three isolates grown in medium used in preparation of inoculum con- tained in shaken and non-shaken flasks were combined before testing. Culture media were filtered three times with suction through Whatman No. 1 filter paper on a Buchner funnel. Filtrates were concentrated lg ygggg to 0.1 original volume at 3200, an equal volume of methanol was added, then the concentrated filtrate and methanol were stored at O-lOC for 40 hr. The precipitate which formed was discarded and methanol was removed from the remaining filtrate in ygggg at 3200. Five ml of the remaining crude concentrate was further purified by gel filtration. Ten g dry spherical polyacrylamide (Bio Gel P-2, 200— 400 mesh) was hydrated by soaking in water 20 hr at 2200. The gel was periodically stirred and water was replaced several times. After packing a K 15/30 Sephadex column, the water was brought even with the top of the gel bed. Five ml of the crude culture concentrate was carefully placed on tOp of the gel bed, the concentrate was brought 47 even with the tOp of the bed, water was carefully added to the top of the column, and the column was sealed. The column was deve10ped with water at 1-2 ml per min. The effluent was collected in 14 five m1 fractions. Each fraction was diluted 1:10 and 1:50 for testing. Possible toxicity of the fractions was tested on g. americana and Celtis occidentalis L. in two experiments and on U. americana and Acer saccharum Marsh. in the third. Succulent cuttings (10 cm) were collected from large trees and placed in 4 ml of each dilution of each fraction. A water control was maintained. An additional amount of the apprOpriate diluted fraction was added to each test tube to maintain the fluid level. None of the fractions collected from culture filtrates of 9. Elm; caused any distinct symptoms on host or non-host cuttings. Some of the cuttings, including controls, deve10ped curled leaves, but this was usually associated with increased turbidity of the fraction, and could have been due to multi- plication of bacteria in the fraction. Bio Gel P—2 would have excluded the 25,000 MW com- pound isolated by Salemink et a1. (53) which was toxic to cuttings of U. hollandica Mill. Cv. belgica. They were also unable to recover the toxic principle by gel filtration using a gel with an inclusion range of 4000-150,000 MW. DISCUSSION AND CONCLUSIONS Experiments were designed to determine if physiological changes occurred in host tissues early in disease deve10p- ment, before macrosc0pic symptom expression began, and to determine the distribution of the pathogen within the host at the time of these events. If physiological changes occurred in pathogen-free tissue, particularly if the pathogen was distant from the site of change, a toxic substance which was mobile in the transpiration stream might be reSponsible for the changes. Physiological changes which occurred in pathogen-free tissues have been reported with several vascular diseases. Foster (23) found that vein-clearing occurred after inocula- tion with Fusarium wilt of tomato before other symptoms were observed. Davis (17), working with intergeneric grafts and E. oxygporum f. lyCOpersici, found non-host scions reSponded to infection in the stock by necrosis. Keyworth (37, 38) worked with multiple grafts between Fusarium wilt—resistant and susceptible tomato plants and found discoloration in pathogen-free resistant tissues. Collins and Scheffer (l6) worked with the same disease, and found respiration and permeability changes in pathogen-free leaves. Scheffer (55) however, did not believe that respiration changes had a major role in disease develOpment because disease could deve10p normally with essentially no change in oxygen uptake. 48 49 Mathre (39) found that photosynthesis decreased in pathogen- free blades of Verticillium-infected cotton plants. Roberts (52) demonstrated an early increase in tranSpiration in both susceptible and resistant elms following inoculation with Q. 31ml. The present work demonstrated that changes in loss of electrolytes and oxygen uptake occurred in pathogen- free tissues of Q. americana in response to Q. Elfli infection. Oxygen uptake by leaf discs from inoculated plants increased to as much as 180 percent of controls 11-22 days after in- oculation. Conductance of aqueous leachates of leaves in- creased to approximately 150 percent of controls in ll-l3 days after inoculation. After reaching a maximum, oxygen uptake decreased to 30-50 percent of controls. A decrease in oxygen uptake was accompanied by a rapid increase in con- ductivity. Leaf chlorosis became evident during this phase of disease development, and indicates that general disruption of host metabolism occurred. These changes are interpreted as occurring in response to some substance which is tranSported through the transpiration stream to the site of action in terminal leaves. However, these data do not prove the in- volvement of a toxin in disease develOpment. This can only be established by isolation and purification of the substance and demonstration that it has a role in disease develOpment. One alternative, that water stress was the cause of the observed changes, was tested and was found not to be involved. 50 Oxygen uptake increased after inoculation with Q. 21mg, but decreased in healthy plants subjected to water stress. There are, however, other factors which could elicit a host re- sponse in non-infected tissues. The observation that oxygen uptake and conductivity changes occur soon after inoculation, before any morphological or anatomical symptoms occur seems to imply that these are early effects in the disease syndrome, but it does not imply that these are necessary for disease development. Conductivity and oxygen uptake changes may indicate that general cell disorganization is occurring. Observations on spore movement and subsequent disease development in American elm have often been made on plants which received enormous numbers of spores, often in large volumes of fluid (2, 4, 20, 49, 52, 64). Comparison of these results with natural infection has been criticized (12, 41). The data reported herein substantiate these criticisms. Q. Elm; was not isolated from twigs of small trees in less than 11 days after inoculation when the trees were inoculated with approximately 400 spores; nor was the fungus isolated from leaf blades and only rarely (1.6 percent) from petioles in the 30 day test period. Comparable (symptomless) leaf blades and petioles from naturally infected trees contained the pathogen in 0.6 percent and 4.5 percent of the attempts. By contrast, 8.0 percent of the leaf blades and 11.5 percent of the petioles from trees which were inoculated with a large spore load contained Q. ulmi. The number of spores herein 51 considered to be a large spore load (5 X 10LI Spores per tree) was much smaller than those used in other colonization studies (2, 4, 49). Experiments which demonstrated rapid distribution of Spores throughout trees and into leaves within 24 hours of inoculation (49) are not representative of natural colonization. Vessels in twigs and in petioles and midveins of leaves on inoculated trees were not visibly occluded when the leaves were taking up oxygen at maximum rate. Sectioning and prepara- tion of tissues for micros00pic observation could have removed occluding gums. However, control of Dutch elm disease can be obtained by chemically-induced occlusion of vessels (59). Control was obtained by preventing Spore movement in the vessels and thus preventing systemic infection. To understand how occlusion can be responsible for both symptom develOpment and resistance, especially since vessel occlusion is complete in resistant-treated trees, is difficult. Most previous work has been done on that phase of the disease during which visible symptoms occur or with tissue containing the pathogen. Leaf symptoms occur concurrently with decreases in oxygen uptake and large increases in loss of electrolytes, and on this basis is rather late in disease develOpment. Many changes have probably occurred in the host by the time leaf symptoms occur. Leaf symptoms may be the manifestation of general cellular disorganization: in this event, all physiological changes which led to disorganization 52 of host cells would have preceeded symptom expression. Therefore, future research should be concentrated on the early events in disease deve10pment, those which occur before macrosc0pic symptoms are observed. l. 7. 9. 10. 11. LITERATURE CITED Banfield, W. M.. and A. L. Smith. 1936. Growth and distribution of Ceratostomella ulmi in tissues of elm. PhytOpathology 25386. (Abstr.) . 1941. Distribution by the sap stream of spores of three fungi that induce vascular wilt diseases of elm. J. Agr. Res. 62:637-681. , E. G. Rex, and C. May. 1947. Recurrence of Dutch elm disease in American elms in relation to tree stature. PhytOpathology 37:1-2. (Abstr.) . 1968. Dutch elm disease, recurrence and recovery in American elm. PhytOpathol. Z. 62:21-60. Beattie, R. K. 1933. How the Dutch elm disease reached America. Proc. Ninth Shade Tree Conf. p. 101-105. Beckman, C. H. 1956. Production of pectinase, cellulases, and growth-promoting substance by Ceratostomella ulmi. Phytopathology 46:605-609. , and F. L. Howard. 1957. Chemical inhibi- tion of Springwood deve10pment in relation to infec- tion and symptoms of the Dutch elm disease. Phyto- pathology 47:3. (Abstr.) . 1958. Growth inhibition as a mechanism in Dutch elm disease therapy. PhytOpathology 48: 172-176. . 1966. Cell irritability and localization of vascular infections in plants. PhytOpathology 56 3 821—8214“ s Brener, W. D., and C. H. Beckman. 1968. A mechanism of enhanced resistance to Ceratocystis ulmi in American elms treated with sodium trichlorOphenyl- acetate. PhytOpathology 58:555-561. Brown, M. F., R. J. Campana, and M. S. Rosinski. 1963. Downward movement of conidia of Ceratocystis ulmi from terminal twigs of elm. PhytOpathology 53:871. (AbStro) 53 12. 13. 14. 15. 16. 17. l8. 19. 20. 21. 22. 23. 54 Campana. F. F. 1967. Spore movement and limited symptom development of Dutch elm disease in small elm stems. Midwestern Chapter, International Shade Tree Conf. Proc. 22:47-52. , and H. G. Brotzman. 1967. Restricted de- ve10pment of Dutch elm disease in small elm stems. PhytOpathology 57:338-339. (Abstr.) Carefoot, G. L., and E. R. Sprott. 1967. Dutch elm disease, p. 198-201. in G. L. Carefoot and E. R. Sprott, Famine on the wind. Rand McNally and Co., Chicago.‘ Collins, C. W. 1941. Studies of elm insects associated with Dutch elm disease fungus. J. Econ. Entomol. 343369- 372. Collins, R. P., and R. P. Scheffer. 1958. Respiratory responses and systemic effects in Fusarium-infected tomato plants. PhytOpathology 48:349- 355. Davis, D. 1954. The use of intergeneric grafts to demonstrate toxins in the Fusarium wilt disease of tomato. Amer. J. Bot. 41:395-398. Dimond, A. E. 1947. Symptoms of Dutch elm disease reproduced by toxins of Graphium ulmi in culture. PhytOpathology 37:7. (Abstr.) Edgington, L. V. 1963. A chemical that retards de- ve10pment of Dutch elm disease. PhytOpatholcgy 53:349. (Abstr.) Elgersma, D. M. 1967. Factors determining resistance of elms to Ceratocystis ulmi. PhytOpathology 572 641—642. Feldman, A. W., N. E. Caroselli, and F. L. Howard. 1949. Physiological and chemotherapeutic investiga- tions of Ceratostomella ulmi. PhytOpathology 39: 6-7. (Abstr. ) . 1950. Physiology of toxin production by Ceratostomella ulmi. Phyto» pathology 40:341- 354. Foster, R. E. 1946. The first symptom of tomato Fusarium wilt: clearing of the ultimate veinlets in the leaf. PhytOpathology 36:69lw694. 24. 25. 26. 27. 28. 29. 30- 31. 32. 33. 34. 35- 36. 55 Frederick, L., and F. L. Howard. 1951. Comparative physiology and pathogenicity of eight isolates of Ceratostomella ulmi. PhytOpathology 41:12-13. (Abs-tr. Gagnon, C. 1967. Hist0chemica1 studies on the alteration of lignin and pectic substances in white elm in— fgcted by Ceratocystis ulmi. Can. J. Bot. 45:1619- 1 23. . 1967. Polyphenols and discoloration in the elm disease investigated by histochemical tech- n1que., Can. J. Bot. 45:2119-2124. Guilkey, P. C. 1957. Silvical characteristics of American elm. U. S. Forest Service Lake States Forest Exp. Sta. Paper 54. 19 p. Guyot, M. 1921. 'Notes de pathologie vegetale', Bull. Soc. Pathol. vegetale France 8:132-136. Abstract in Rev. Appl. Mycology 1:334-335. Hart, J. H. 1960. A modified method for culturing elm samples. Plant Dis. Rep. 44:806—807. , W. E. Wallner, M. R. Caris, and G. K. Dennis. 1967. Increase in Dutch elm disease associated with summer trimming. Plant Dis. Rep. 51:476-479. . 1968. Personal communication. Holmes, F. W. 1964. Virulence in Ceratocystis ulmi (Buisman) C. Moreau. PhytOpathology 54:128. TAbstr.) . 1965. A test clone of Ulmus americana uniformely susceptible to Ceratocystis ulmi. PhytOpathology 55:1284. (Abstr.7— . 1967. Resistance of certain elm clones to Ceratocystis ulmi and Verticillium albo-atrum. Phytopathology 57:1247-1249. Kais, A. G., E. B. Smalley, and A. J. Riker. 1962. Environment and deve10pment of Dutch elm disease. PhytOpathology 52:1191-1196. Kerling, L. C. P. 1955. Reactions of elm wood to attacks of Ophiostoma ulmi (Buism.) Nannf. Acta Bot. Neerl._4}3984463. 37- 38. 39- 40. 41. 42. 43. an. 45. 46. 47. 56 Keyworth, W. G. 1963. The reaction of monogenic re- sistant and susceptible varieties of tomato to in- oculations with Fusarium oxysporum f. 1ycopersici into stems or through Bonny Best root stocks. Ann. Appl. Biol. 52:257-270. . 1964. Hypersensitivity of monogenic re— sistant tomato scions to toxins produced by Bonny Best rootstocks invaded by Fusarium oxysporum f. lchpersici. Ann. Appl. Biol. 54:99-105. Mathre, D. E. 1968. Photosynthetic activities of cotton plants infected with Verticillium albo—atrum. PhytOpathology 58:137-141. May, 0. 1930. Dutch elm disease in Ohio. Science 72:142—143. Neely, D. 1968. Twig inoculations on American elm with Ceratocystis ulmi. PhytOpathology 58:1566-1570. Northen, H. T. 1943. ,Relationship of dissociation of cellular proteins by incipient drought to physio- logical processes. Bot. Gaz. 104:48-485. 0ue11ette, G. B. 1962. Morphological character- istics of Ceratocystis ulmi (Buism.) C. Moreau in American elm trees. Can. J. Bot. 40:1463-1466. . 1962. Studies on the infection process of Ceratocystis ulmi (Buism.) C. Moreau in American elm trees. Can. J. Bot. 40:1567-1575. Parker, K. G., P. A. Readio, L. J. Tyler, and D. L. Collins. 1941. Transmission of the Dutch elm disease pathogen by Scolytus multistriatus and the ggvelopment of infection. PhytOpathology 31:657- 3. , D. L. Collins, L. J. Tyler, D. P. Connola, W. E. Ozard, and H. Dietrich. 1948. The Dutch elm disease: association of Ceratostomella ulmi with Scolytus multistriatus, its advance into new areas, methods of determining its distribution, and control fig the disease. Mem. Cornell Agr. Exp. Sta. 275. p0 Pomerleau, R. 1965. The period of susceptibility of Ulmus americana to Ceratocystis ulmi under conditions prevailing in Quebec. Can. J. Bot. 43:787-792. 48. 49. 50. 51. 52- 53- 54. 55- 56. 57- 58- 59- 57 . 1965. Seasonal transmission of the elm disease by Hylurgopinus rufipes in Quebec. Can. J. Bot. 43:1592-1595. , and A. R. Mehran. 1966. Distribution of spores of Ceratocystis ulmi labelled with Phosphorus-32 in green shoots and leaves of Ulmus americana.Le Natur. Can. 93:577-582. , and R. Pelletier. 1967. A new technique to determine the presence of Ceratocystis ulmi. (Buism.) C. Moreau in fresh elm leaves, shoots, and wood. Le Natur. Can. 94:59—62. Pringle, R. B. and A. 0. Braun. 1957. The isolation of the toxin of Helminthosporium victoriae. Phyto- pathology 47:369-371. Roberts, B. R. 1966. Transpiration of elm seedlings as influenced by inoculation with Ceratocystis ulmi. Forest Sci. 12:44-47. Salemink, C. A., H. Rebel, L. C. P. Kerling, and V. Tchernoff. 1965. Phytotoxin isolated from liquid cultures of Ceratocystis ulmi. Science 149:202-203. Scheffer, R. P. and J. C. Walker. 1953. The physiology of Fusarium wilt of tomato. PhytOpathology 43: 116-125. . 1960. Variations in respiratory responses in Fusarium-infected tomato plants. PhytOpathology 503192-1950 Schneider, G. W., and N. F. Childers. 1941. Influence of soil moisture on photosynthesis, respiration, and transpiration of apple leaves. Plant Physiol. 168565-5830 Schreiber, L. R. 1966. Survival of Ceratocystis ulmi in American elm seedlings. PhytOpathology 56:899. (Abstr.) , and R. J. Stipes. 1967. The effect of in- oculum spore concentration on the development of foliar symptoms of Dutch elm disease. PhytOpath- ology 57:1269. Smalley, E. B. 1962. Prevention of Dutch elm disease by treatments with 2,3,6-trichloropheny1 acetic acid. PhytOpathology 52:1090-1091. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 58 . 1963. Seasonal fluctuations in susceptibility of young elm seedlings to Dutch elm disease. Phyto- pathology 53:846-853. Smucker, S. J. 1940. Apparent recovery of American elms inoculated with Ceratostomella ulmi. PhytOpathology 30:1052-1054. Swingle, R. U. and R. R. Whitten. 1967. Dutch elm_ _ disease, p. 177-180. lg Davidson, A. G. and R. M. Prentice (ed.) Important forest insects and diseases of mutual concern to Canada, The United States, and Mexico. R. Duhamel, F. R. S. C. Queen's printer and controller of stationary. Ottawa, Canada. Tchernoff, V. 1963. Vegetative propagation of elms by means of shoots out from callused roots. Acta Bot. Neerlo 12 840-50 0 Tyler, L. J., and K. G. Parker. 1945. Pathogenicity of thg Dutch elm disease fungus. PhytOpathology 35: 257-2 1. . 1945. Influence of temperature on Dutch elm disease in potted American elm. Phytopathology 35: 302-304. Umbreit, W. W., R. H. Burris, and J. F. Stauffer. 1964. Manometric techniques. 4th ed. Burgess Publ. Co., Minneapolis, 305 p. Verall, A. F., and T. W. Graham. 1935. The transmission of Ceratostomella ulmi through root grafts. Phyto— pathology 25:1039-1040. Whitten, R. R., and W. A. Reeks. 1967. Smaller EurOpean elm bark beetle Scolytus multistriatus Marsham, p. 167- 170. lg Davidson, A. G., and R. M. Prentice (ed.) Important forest insects and diseases of mutual con- cern to Canada, The United States, and Mexico. R. Duhamel, F. R. S. C. Queen's printer and controller of stationary. Ottawa, Canada. Wilson, 0. L. 1965. Ceratocystis ulmi in elm wood. Phytopathology 55:477. Zentmyer, G. A., and J. S. Horsfall. 1942. Toxin forma- tion by Ceratostomella ulmi. Science 95:512-513. , and P. P. Wallace. 1944. Factors influencing the deve10pment of Ceratostomella ulmi in elm trees. PhytOpathology 34:1014. (Abstriy 72. 59 , J. G. Horsfall, and P. P. Wallace. Dutch elm disease and its chemotherapy. Exp. Sta. Bull. 498. 70 p. 1946. Conn. Agr. "WMMW“ ‘