FUNDAMENTALS FOR TAXONOMIC STUDIES OF FUSARIUM1

By H. W. WOLLEN WEBER,2 Pathologist, Biologische Reichsanstalt für Land- und Forstwirtschaft; C. D. SHERBAKOFF,3 Pathologist, Agricultural Experiment Station, University of Tennessee; and O. A. REINKING, Pathologist, United Fruit Company; with the cooperation of HELEN JOHANN, Assistant Pathologist, Office of Cereal Investigations, Bureau of Plant Industry, and ALICE A. BAILEY, Junior Pathologist, Office of Cotton, Truck, and Forage Crop Disease Investiga- tions, Bureau of Plant Industry, United States Department of Agriculture 4

INTRODUCTION volved the United Fruit Co.'s bringing Dr. H. W. Wollenweber over from Ger- There has been a long-felt need for many, and Dr. O. A. Reinking from a general conference of all workers on Central America. The United States the Fusarium problem in order that a workers included Dr. C. D. Sherbakoff, uniform taxonomy may be formulated of the University of Tennessee, Miss upon which future work can be based. Helen Johann and Mrs. Alice A. Bailey, The various workers have always been representing, respectively, the Offices fundamentally in reasonably close of Cereal Investigations, of Cotton, agreement regarding the principles of Truck, and Forage Crop Disease In- Fusarium classification, but there has vestigations, and of Pathological col- been a number of points which were lections of the United States Depart- not clearly understood, possibly because ment of Agriculture, in a joint study of the investigators have not been suffi- Fusarium classification. The United ciently in contact with each other. Fruit Co., through its director of agri- Variations in the methods employed in cultural research, Dr. John R. John- different laboratories have complicated ston, adopted this liberal policy in the problem of species identification. supporting the scientific work to The fact that the fungi vary according a greater extent than their purely to the environment, age, and source of economic interests required. It was culture and the failure of different agreed, on the invitation of Dr. L. R. workers to understand one another has, Jones, to hold the conference at the it is believed, largely accounted for the University of Wisconsin, in Madison. differences arising from Fusarium The purpose of the conference was studies. to give an opportunity to cooperate in Recent investigations on the banana- a personal way, to compare cultures wilt problem conducted by the United assembled from all possible sources, Fruit Co. in Central America showed and in this manner to clear up the the immediate necessity of a general somewhat tangled taxonomy of this conference on the Fusarium problem. difficult genus. All the important Because of the large number of related European cultures, those being studied Fusaria encountered in a study of the at present in the United States, and banana - wilt - producing organism on the collections made from the tropics, banana plantations, it was deemed ad- primarily from Central America, were visable to complete the fungous inves- assembled at the meeting for special tigations with a final study in Europe study and comparison. The work of or in the United States. The proposed the conference covered, in so far as studies of the United Fruit Co. were possible, the study, comparison, and discussed at the meeting of scientific identification of the specimens and societies in Cincinnati, in December, cultures of the fungi then available. 1923, and through Dr. W. A. Orton a The main studies were made on the plan for an American conference on tropical collection, since it was agreed, Fusarium was proposed. The idea in- because of the part taken by the

1 Received for publication Sept. 29, 1924; issued June, 1925. 2 Engaged in special investigations for the United Fruit Co. during the Fusarium conference at Mad- ison, Wis., in August, 1924. 3 Present at the conference in cooperation with the Offices of Cotton, Truck, and Forage Crop Disease Investigations, of Cereal Investigations, and of Pathological Collections, Bureau of Plant Industry, U. S. Department of Agriculture. 4 The writers are indebted to the Departments of Plant Pathology and of Botany of the University of Wisconsin for providing facilities with which to carry on the investigations during the period oí the con- ference.

Journal of Agricultural Research, Vol. XXX, No. 9 Washington, D. C. May 1, 1925 Key No. G-469 19976—25t 4 (833) 834 Journal of Agricultural Research Vol. XXX, No. 9

United Fruit Co., that the conference others may be determined directly should facilitate, in the best way from the fungus growing under natural possible, a satisfactory definition of conditions without resorting to pure the numerous species found on banana cultures. Other Fusaria, which appear plantations. The studies made of the primarily in a microconidial stage entire collection embodied species from under natural conditions and produce all sections of the genus Fusarium, only a few sickle-shaped spores, are including important border-line strains, more difficult to identify. For these thereby making it possible to arrive the so-called normal growth must, at at a uniform taxonomy of the group. first, be produced. It is, therefore, It is believed that the main points of necessary to make a careful study from difference have been agreed upon and pure culture in order accurately to that the present paper will present a identify these organisms. Generally, clearer understanding of the Fusarium macroconidia are regarded as the problem from the standpoint of the normal spore type. In certain groups, identification of species. however, the microconidia may have Material of each species studied and definite characters, such as a pearlike identified at the conference will be shape or a formation in chains, which deposited in the pathological collec- aid in the determination of the section tions herbarium of the Bureau of (Sporotrichiella, Arthrosporiella) and, Plant Industry, United States Depart- in exceptional cases, may even lead to ment of Agriculture, for permanent the identification of species (F. poae preservation. The cultures will be (Peck) Wr., F. moniliforme Sheld., F. prepared in the manner most satis- decemcellulare Brick.). factory for the preservation of her- Other normal reproductive stages, barium material. Pure cultures of such as chlamydospores, by their each of the species studied and iden- presence or absence indicate the border tified will also be placed in the bureau. of certain groups (Elegans, Lateritium). These cultures will have substantially Sclerotia may be characteristic for the status of type specimens. groups (Lateritium) and even for a number of species {F. sclerotium Wr.). CRITERIA OF THE NORM Color of the conidia and color of the aerial mycelium and stroma are further Only a résumé of the criteria of the reliable characters for taxonomy. norm will be here considered, since a Minor characters such as mycelium, complete discussion of the subject has hyphae with specialized cells, coremia- appeared in former publications (5, 9).b like aggregations of hyphae, and aro- The normal condition of Fusaria from matic odors may also be of some the standpoint of determination may importance for differentiation. In op- be present in nature, but generally it position to the norm there are abnormal must be produced by growing the features which might be valuable to fungi under pure culture conditions. mention in the diagnosis. Not so A few Fusaria such as F. dimerum much stress, however, is placed on Penz., F. scirpi Lamb, et Fautr., F. degenerated conditions in the life culmorum (W. G. Sm.) Sacc, and some cycle of Fusaria.

5 Reference is made by number (italic) to "Literature cited," p. 843. May 1, 1925 Taxonomic Studies of Fusarium 835

-Stages in the life cycle of Fusaria: A-F, H and S, microconidia. I-R, T-X, macroconidia. Y, Swellings of hyphae. Z, Ai-Di, Chlamydospores. Ei-Fi, Conidial propagation from conidia. Gi, Sclerotium. Hi, Sporodochium. Ii, Sclerotiain clusters on an erumpent base. Ki, Columns formed by conidia borne in chains. Li,-Ni, Pionnotes covering either the medium directly (Li), or a mycelial sheet (Mi), or numerous associated sporodochia (Ni) formed on the substratum. Examples: Fusarium poae (Peck) Wr. (A, J), F. orthoceras App. et Wr. (B, E, K, Ai, Bi), F. moniliforme Sheld. (C, F, G, GO, F. chenopodinum (Thüm.) Sacc. (D, M, N, Z), F. argillaceum (Fr.) Sacc. (H, I, Ci, DO, F. subcarneum Crouan (L), F. flavum (Fr.) Wr. (O), F. dimerum Penz. (P), F. salicis Fuck. (Q), F. sarcochroum (Desm.) Sacc. (R, S), F. vitícola Thüm. (T), F. solani (Mart. pr. p.) App. et Wr. (U), F. vasinfectum Atk. (V, normal, W, swollen, X, dried spore), F. tricinctum (Cda.) Sacc. with swellings of hyphae (y), F. macroxysporum Lindf. with pionnotal propagation of macroconidia (Ei) and development of microconidia (Fi). 836 Journal of Agricultural Research Vol. XXX, No. 9

For a proper understanding of the tions except Ventricosum terms (8) used for normal forms of re- and some species of Ele- productive stages and mycelium of gans and Eupionnotes Fusaria the characters considered of (#) in columns represent- primary importance are enumerated ing a chainlike develop- below. ment, as in F. stilboides I. NORMAL REPRODUCTIVE STAGES Wr. (fig. 1, KO. c. Shape: Fusoid-lanceolate, 1. Microconidia: 0-(0-3)-septate7 globose-pearshaped as in F. slightly curved, mostly poae (Peck) Wr. (fig. 1A); el- apedicellate, F. argilla- lipsoidal as in F. orthoceras ceum (Fr.) Sacc. (fig. 1, App. et Wr. (fig. 1, B); ovoid- I); slightly sickle-shaped, fusoid as in F. moniliforme apedicellate, F. betae Sheld. (fig. 1, C); comma- (Desm.) Sacc. (fig. 1, K) ; shaped as in F. chenopodinum sickle-shaped, pedicellate, most of the species; (Thüm.) Sacc. (fig. 1, D). strongly dorsiventral, F. They may be found scattered scirpi Lamb, et Fautr. (fig. 1, A-D), in false heads and F. chenopodinum (fig. 1, E), or in chains (fig. 1, (Thüm.) Sacc. (fig. 1, N) ; F) and may form sporodochia terminal cell suddenly and pionnotes. constricted, F. culmorum 2. Macroconidia: O-pluriseptate, (W. G. Sm.) Sacc. and dorsiventral, fusoid to sickle- F. chenopodinum (Thüm.) shaped, apedicellate and pedi- Sacc. (fig. 1, M); terminal cellate, scattered, sporodoch- cell whiplike, much elon- ial and pionnotal. gated F. scirpi Lamb, et a. Septation types: Fautr. and F. chenopo- 0-septate, F. flavum dinum (Thüm.) Sacc. (fig. (FT.) Wr. (fig. 1,0); 1, N); curvature within 1-septate, F. dimerum the limits of conical sec- Penz. (fig. 1, P); tions as elliptic (fig. 1, 3-septate, apedicellate, M), parabolic and hyper- F. betae (Desm.) Sacc. bolic (fig. 1, N). The (fig. 1, K); curvature is mainly de- 3-septate, pedicellate, F. termined from the dorsal poae (Peck) Wr. and and ventral sides of (fig. l, J); spores, in side view, with- 5-septate, F. chenopodi- in the septation regior. num (Thüm.) Sacc. 3. Chlamydospores: Terminal (fig. 1, and (fig. 1, M); Bi, Di), intercalary (fig. 1, Ai), 5-7-septate, F. cheno- single, in chains, in cluster?, podinum (Thüm.) mvcelial (fig. 1, Z to Bi), coni- Sacc. and (fig. 1, N) ; dial (fig. 1, Ci). 9-septate, F. decemcel- a. Terminal: In sections Ele- lulare Brick. gans, Martiella, Ventri- b. Sporulation types: Scat- cosum; unicellular, F. ar- tered; in false heads; gillaceum (Fr.) Sacc. (fig. pionnotes, continuous, 1,N Di); 0-1-septate, sec- gelatinous, slimy, (1) tions Martiella and Ele- formed on the naked gans (fig. 1, Bi). Most surface of the substratum fungi forming terminal (fig. 1, Li) as in section spores may also develop Eupionnotes, species of intercalary spores. In section Roseum, Gibbo- section Ventricosum no sum, Elegans and Dis- intercalary spores are color, or (#) on a mycelial produced. sheet as in F. avenaceum b. Intercalary: Mostly 0-1-sep- (Fr.) Sacc. (fig. 1, Mi) or tate (fig. 1, Ai) in sections (3) composed of an aggre- Elegans, Martiella, Dis- gation of sporodochia as color, Gibbosum, and in F. martii App. et Wr. Eupionnotes (subsection (fig. 1, Ni); sporodochia, Chlamydospora). Chains (1) tubercularialike (fig. and clusters (fig. 1, Z) 1, Hi) found in the ma- may also occur in most jority of species of all sec- of these sections. Mav 1, 1925 Taxonomic Studies oj Fnsarium 837

4. Sclerotia: Plectenchymatic struc- characteristic colors produced tures, general appearance on sterilized potato tubers similar to perithecia of Gib- are carmine red by species berella, but with uniform of section Discolor, Roseum, structure throughout. Glo- Sporotrichiella, and some of bose or rugose, single or in Gibbosum, and citric or sul- cauliflowerlike clusters (fig. 1, phuric yellow in Neesiola Gi). Color, blue in sections subsection of Discolor. These Roseum, Gibbosum, Lateri- colors are of the basic modi- tium, Discolor, Elegans, and fication and consequently are Martiella; brownish to dark not turned blue by the addi- brown in Arthrosporiella (F. tion of alkali. The carmine diversisporum Sherb.), Gib- red turns yellow with an ad- bosum, and some species of dition of acid. No color con- section Discolor. trast is observed in sections II. COLOR CHARACTERS such as Eupionnotes, Arach- 1. Color of conidia: Brownish white nites, and Ventricosum. The to golden yellow in section stroma and aerial mycelium Martiella; brownish white to of these fungi have the color pale orange in section Elegans, shades of their conidia. The Discolor, Gibbosum, and Ar- color of sclerotia and sclerotial throsporiella; orange in sec- stroma is blue in some and tion Roseum Lateritium, Eu- brown in other groups, as dis- pionnotes and Arachonites. cussed under sclerotia. The color of conidia is suffi- III. MINOR CHARACTERS ciently constant to be a reli- 1. Hyphae with swollen cells: Not able character for some true chlamydospores (fig. 1 Y) groups. as in F. flocciferum Cda. and 2. Color of aerial mycelium: White, F. tricinctum (Cda.) Sacc. rose to yellow and blue, 2. Aerial mycelium: Loose, dense, mostly representing a diffused jellylike, cottony, radiate, color of the stroma. Changes zonate. in color of mycelium due to 3. Immersed mycelium: Slimy or the reaction of the sub- leathery sheet (stroma), that stratum is substantially the may or may not have a same as that discussed under plectenchymic base. Section color of stroma. Eupionnotes. 3. Color of stroma: Brownish white, carmine, yellow, and blue. 4. Coremialike aggregations of hyphae. The acid modification is espe- Aggregations of anastomosed cially well developed on ster- hyphae as in Eupionnotes and ilized rice and may aid in the Ventricosum. True coremia identification of Fusarium so far were not observed. groups. This acid modifica- 5. Aromatic odor: Somewhat similar tion differs in some sections to lilac odor, produced by a for it is golden yellow in most number of species of section of the species in section Dis- Elegans such as F. oxysporum color, Saubinetii, Roseum, Schlecht., F. hyperoxysporum Sporotrichiella, and in some Wr., F. zonatum (Sherb. s. of Gibbosum; rose to vine red var.) Wr., F. vasinfectum Atk., in many species of section and F. cúbense Erw. F. Sm. Elegans; and rose to brown The last three species have color, with a tendency toward the strongest odor, especially diffusing into rice, in section when grown on rice. Martiella. The basic color IV. ABNORMAL CHARACTERS modification is mostly blue or 1. Spore: Exceptional size, irregular- violet. In general, all of these ities in shape of conidia, fungi produce on rice an acid swollen cells in germinating color modification that may spores (fig. 1, W) and con- change gradually with age stricted cells in dry conidia into an alkaline modification. (fig. 1, X). These abnormal This color change can be characters, although frequent- produced at once by the ly present, are not taken into addition of sufficient alkali consideration for taxonomic to a well-developed culture of purposes and have been more the fungus on rice. Other fully discussed elsewhere (/). Journal of Agricultural Research

2. Mycelium: With the exception of nected with one and the same fungus, Eupionnotes, in which im- F. blackmani Brown et Home. mersed growth of the myce- This variability of the mycelium and lium predominates even under spore character has been found to normal and favorable spore occur with various fungi. Notable production, changes to an among these are two closely related immersed gelatinous growth Fusaria described as F. anguioides in general indicate degener- Sherb. and F. anguioides var. caudatum ation and self-digestion ac- Sherb. (5). These two organisms are companied by abnormal spore closely related to F. anthophilum (A. production or even sterility. Br.) Wr. isolated from apple fruit in This condition maybe changed England. The abnormal variations into the norm by transferring found in these cases, though great, to various metfia or by selec- were not mentioned in the diagnosis tion of virulent aerial colonies of the Fusaria because normal charac- from new agar plates. For- ters were regarded as sufficient for de- tunately the number of Fu- termination. saria showing a tendency to degenerate seems to be small. This tendency is notable in PEODUCTION OF THE NORM Fusarium nivale (Fr.) Sacc, F. anthophilum (A. Br.) Wr., Cultural conditionswhich will produce F. orthoceras App. et Wr., good mycelial growth are not necessarily and F. flocciferum Cda. The those most favorable for the production latter two fungi if inoculated of normal spores. In the study of Fusar- into living potato tubers, ium it has been found necessary to thoroughly disinfected with use a variety of media for the pro- formaldehyde, regain their duction of the various normal mycelial normal behavior when re- and reproductive' stages. The best isolated and retransferred to media for these purposes at present sterilized media. In some available are the ordinary ones used in . cases, agar media too rich in plant pathology investigatio sugar or too alkaline or acid combination of a number of vi w in reaction favor degeneration. substances may prove valuable Conditions, however, that fa- some cases, but, so far, sufficient work vornormal conidial production has not been done to Justify recom- generally warrant longevity mending such a medium. Because the and sufficient constancy. On same vegetable under different con- the same medium mycelium ditions varies in its chemical compo- transfers often give only a ster- position, it would be highly desirable ile stroma, while a transfer of to produce a satisfactory synthetic macroconidia increases the medium. It is hoped that a synthetic tendency to reproduce this nutrient may sometime be found stage. These facts if not which will render it possible to control fully understood may make the production of spore bodies or the proper identification of cer- sclerotia at will. Until some such new tain species doubtful. They media are produced, vegetable media show the necessity of im- will be found to be fairly satisfactory proving cultural methods. for taxonomic purposes. Brown and Home (?) in their studies No one culture medium is at present of the genus Fusarium have given used which will produce optimum interesting details on the modifying development of all phases of fungus effect of transfers from various parts growth; hence certain media have been of a given fungus. These authors have selected which proved most satisfac- stated that the rate of spread of certain tory for the best production of conidia, colonies fell off after a time and was and other media for the production of reduced (staling), while other colonies sclerotia, mycelium, chlamydospores, developed into normal sporodochia or and other characters. It must be pionnotes with small and low septate understood, however, that not all or long and high septate conidia. Fur- species will react in exactly the same thermore, "saltants" showing sectorial way, and therefore a medium which effects of some strains have been pro- produces normal conidia in most duced in Fusaria derived from six dif- species may need to be supplanted by ferent isolations from apple. In the another in the case of other species. preliminary studies this wide range of The media used at the conference are variability was considered to be con- given in Table I. May 1, 1925 Taxonomic Studies of Fnsarium 839

a TABLE I.—Culture media showing types of growth readily produced on each.

Media conidia | Älsclmtiai S" Color

Potato tuber cylinders (no water added) + + + +(basic) Oatmeal agar (Sherbakoff (5) ) + + ¡ -(-(basic) Potato agar+2 per cent dextrose (200 gm. to + + ! 1,000 c. c.) Potato agar+5 per cent dextrose (200 gm. to + ' + -(-(basic) 1,000 c. c.) Rice (2gm.to(5c. c. water) +----' +(acid) Melilotus stems (mature stems+4 c. c. water)... + + Color of spores Lupinus stems (mature stems+4 c. c. water) + +.-- Color of spores Alnus (2 to 3 year twigs+4 c. c. water) + + ¡ Color of spores

0 All agar media contains 2 per cent agar. All media are autoclaved for 45 minutes at 10 pounds pressure except rice, which is steamed for one hour on three successive days. When necessary to use young Melilotus stems, 4 c. c. of a 0.5 per cent KOH solution should be used in place of water to neutralize the acidity of the stems. In general, the use of several vegeta- consideration are temperature, humid- ble media, such as those mentioned ity, and light. There are indications, in the table, will result in the pro- in some species at least, that tempera- duction in good condition of the ture has an appreciable effect upon the various characteristics of the fungus. morphology of the conidia as well as In some cases special methods must be upon the rate of growth of mycelium employed for the production of certain (4)- Perhaps to temperature may also phases. The addition of acid to the be ascribed part of the effect on color standard potato agar greatly stimulates intensity usually attributed to light. an abundant porduction of normal Humidity affects the nature of the spores in such species as the Fusarium culture, as is evidenced by the increased stage of Gibberella saubinetii (Mont.) production of chlamydospores in some Sacc. Substances poor in food value species and the production of swollen favor the production of chlamydospores. conidia in excessively moist media. In species which produce chlamydo- The fact that these environmental spores only rarely, as F. aurantiacium factors have been given very little con- (Lk.) Sacc, chlamydospore develop- sideration in identification of Fusaria ment may be brought about by may be responsible for some of the growing in sterile tap water. The difficulties encountered in various coun- reaction of the medium is important tries where taxonomic work is done. from the standpoint of color production and, in some cases, greatly influences IDENTIFICATION growth of mycelium and abundance of spore production. If we realize the fact that tubercular In transferring, it must be borne in sporodochia with normal and uniform mind that the type of inoculum conidia occur in the majority of Fusaria influences the resultant growth. Con- that can be easily grown in pure cul- tinuous transfer of one type of in- ture, there will be no difficulty in oculum tends to production of that judging the normal stages and, con- type of growth. Repeated transfer sequently, in identifying most of these of mycelium or chlamydospores tends fungi. For the remaining forms more to good development of mycelium and detailed studies will have to be con- sclerotia in those species which have ducted in order to produce or prove sclerotia. For production of abundant the absence of particular stages such spores, transfers should be made as sporodochia, pionnotes, chlamydo- from sporodochia or pionnotes. When spores, and sclerotia. It should be sporodochia are lacking and macro- stated also that even conidia normal spores are not numerous, the chance of in appearance often differ greatly in transferring mycelium may be elimi- their size and shape according to where nated by making dilutions in tubes of they are produced, whether on mycel- sterile water and transferring a loop ium' over a wet surface or in a definite full of the dilution to the culture sporodochium. In a typical pion- medium, or by using the plate dilution notes long conidia may be produced, method. The latter method affords and in sporodochia they may be con- an opportunity to select colonies siderably shorter, as in F. vasinfectum showing the greatest tendency toward Atk., F. cúbense Erw. F. Sm., F. spore production. lycopersici (Sacc.) Wr., and others. Environmental factors other than Differences in size and shape may also the medium which should be taken into appear when the fungus is grown on 840 Journal of Agricultural Research Vol. XXX, No. 9

various types of media. In section sections. The color of conidia and Roseum longer and slenderer conidia stroma especially represent an impor- are produced on rice and oatmeal agar tant complex for identification. than on other media more favorable The ordinary procedure in determin- to sporodochial development. Because ing Fusaria is to illustrate and measure of these differences, it is necessary normal spores and a few exceptions that to pay particular attention to the show the variability and changes with average size of different types as grown increasing age. The standard magnifi- upon different media and under differ- cation used for the drawings is 1-1000 ent conditions. A comparison must for the spores and 1-500 for parts of be made only of conidia produced under sporodochia showing conidiophores and comparable conditions and between formation of spores. Smaller magnifi- comparable sporulation types. cations may be used for special pur- Often the identification is difficult poses, such as groups of sporodochia, because certain species persistently sclerotia, and stroma erumpent. For produce only a microconidial stage. a short study water mounts are prefer- In such instances some special methods able. In order to prevent moisture must be employed to induce the from escaping during the procedure of fungus to develop macroconidia. For drawing, the edges of the cover glass instance, in the case of F. chenopodinum may be waxed. Another method (7) of (Thum.) Sacc. only the comma-shaped preparing slides, which has the ad- spore (fig. I, D) may be observed for vantage of placing all the spores in a long time. However, by repeated one plane and holding them in place, transfers of occasionally found macro- is to make the mount on a very thin conidia, finally normal macroconidia (0.5 mm.) sheet of agar placed on the can be obtained in abundance. These slide. These agar sheets can readily conidia enable us to identify the fungus be prepared by pouring clear agar with ease. It is important to empha- between glass slides set the required size that, in this species as in some distance apart. This method is very others, often an intermediate type of convenient for photographic purposes. macroconidia is produced when the fungus is in a semimycelial stage of growth and when sporodochia are CLASSIFICATION OF THE FORM- underdeveloped. These conidia in F. GENUS FITSARIUM chenopodinum (Thum.) Sacc. can easily be taken for those of F. sambucinum The form-genus Fusarium includes Fuck. Generally the microconidia in all hyphomycetes and conidial stages of themselves do not represent definite ascomycetes that have no black or enough characters to be used for iden- pure gray color either in mycelium or tification. in conidia and that have macroconidia In those doubtful cases wThere nor- that are acrogenous, typically septate, mal macroconidia are not readily sickle-shaped, and not round at the produced the presence of chlamy- ends. Microconidia, chlamydospores, dospores may be of extreme import- and sclerotia may be present. Some ance for determining the group. The Fusaria have been proved to be conidial presence or absence of chlamydospores stages of certain Hypocreaceae, such as makes it possible to separate species of Nectria, Hypomyces, Gibberella, and section Elegans with terminal chlamy- Calonectria. The fungi of the form- dospores from section Lateritium genus Cylindrocarpon, formerly con- with no chlamydospores, even though sidered as Fusaria, have Nectrias and their macroconidia may be similar certain Hypomyces {8) as perfect stages. when grown under certain cultural Some of the species of Ramulara or conditions. similar fungi as Septomyxa also form- The presence of ascigerous stages in erly placed under Fusarium, have as some of the sections is of additional their perfect stage Neonectria {10, 11) help in the identification of the imper- and Mycosphaerella {11). The fungi fect or conidial forms (9, 11). A num- parasitic on scale insects and possessing ber of ascomycetes (Nectria, Calonec- sickle-shaped to fusoid conidia are re- tria, Hypomyces, Gibberella) have ferred to the genus Microcera, wiiich been grown from the ascospores in pure is said to be the conidial stage of culture and developed the conidial Sphaerostilbe. stage, thus showing the relationship. In other cases cultures of Fusaria col- GROUPING OF FUSARIA IN SECTIONS lected from nature have developed perithecia. Many Fusaria readily fall into sepa- Color characters on various media rate groups possessing similar charac- rich in carbohydrates afford a fairly ters. These groups are called sections, reliable means of placing the fungi into which, on the basis of their apparent May 1,1925 Taxonomic Studies of Fusarimn 841 relationship, especially when their per- Wr.; 8, Liseola n. n.6; 9, Lateritium feet stages are considered, are arranged Wr.; 10, Discolor Wr.; 11, Spicarioides as follows: 1, Eupionnotes Wr.; 2, N.comb.7;r2,SaubinetiiWr.; 13,Elegans Arachnites Wr.; 3, Sporotrichiella Wr.; Wr.; 14, Martiella Wr. (sensu extenos)8; 4, Camptospora Wr.; 5, Arthrosporiella and 15, Ventricosum Wr. Sherb.; 6, Gibbosum Wr.; 7, Roseum

KEY TO THE SECTIONS OF FUSARIUM

With the exception of section Camptospora (10), which needs further study, all of the species are arranged in the following key to the sections : a. Microconidia on aerial mycelium usually present and dominately 0-septate, ovoid, fusoid, reniform, or pearshaped. b. 0-septate conida pearshaped. Macroconidia when normal are in shape intermediate between those of the sections Elegans and Roseum, though more curved than either; substratum rose-colored, intercalary chlamydospores, or similar structures may be present. Sec. 3, Sporotrichiella. bb. 0-septate not pearshaped. c. 0-septate conidia in chains. d. Conidial walls thin. Microconidia mostly in chains, fusoid- ovoid; macroconidia in form and color similar to those of sec. Lateritium; no chlamydospores; substratum vina- ceous-violet; some of the species are connected with Gibberellas of sec. Lisea (Sacc.) Wr Sec. 8, Liseola n. n. dd. Conidial walls thick or highly refractable. Macroconidia pluriseptate, in shape resembling those of the sec. Dis- color Sec. 11, Spicarioides n. comb. ce. 0-septate conidia not in chains. d. Conidial walls thin. Macroconidia attenuate at the top ends, pedicellate; terminal and intercalary chlamydo- spores present; color of conidia brownish to salmon; no blue or green color in conidia even as a diffusion from stroma; stroma on artificial media principally vinaceous to lilac Sec. 13, Elegans. dd. Conidial walls relatively thick. Macroconidia somewhat truncate or rounded at the top end, or at least not distinctly attenuate, some times slightly constricted at the tip ends; terminal and intercalary chlamydospores present; color of conidia brown-white to golden brown with occurrence of green to green-blue as diffusion from stroma Sec. 14, Martiella. aa. Microconidia on aerial mycelium usually absent or 0-3 or more septate, reniform, comma, spindle, to sickleshaped. b. Macroconidia apedicellate. Color type, orange to light salmon, c. Typical pionnotes always present; comparatively slow- growing fungi Sec. 1, Eupionnotes. cc. Typical pionnotes absent; comparatively fast-growing fungi Sec. 2, Arachnites. bb. Macroconidia subpedicellate to pedicellate. c. Terminal chlamydospores present; intercalary chlamydo- spores absent; no true sporodochia; macroconidia wedge- shaped to slightly sickle-shaped, not constricted at the top. Sec. 15, Ventricosum.

« Liseola nn. (Syn. Constrictum Wr. pro parte subs. Elegantis; Moniliforme Sherb.b); Microconidiis plus minusve in catenulis, dispositis, fusoideis-ovoideis, macroconidiis forma et colore sectionis Lateritii, liberis, in sporodochiis, in pionnote; chlamydosporis nullis; stromaet violáceo. Status conidicus Gib- berellarum sect. Liseae (Sacc.) 7 Wr. Spicariodes (Wr. subsect.) n. comb. (Syn. subsect Spicarioides sectionis Discoloris Wr.) Micro- conidiis sporodochialibus pluriseptatis forma specierum sectionis Discoloris. Chlamydosporis nullis Stromate carmíneo. s Martiella Wr. (Sensu extenso sectionis Martiellam WTr. et Pseudomartiellam Wr. includendo). Macro- conidiis dorsiventralibus fusoideofalcatis, apice rostrato truncato vel rotundato, basi plus minusve sub- pedicellata, in sporodochiis et pionnote sordide albo, ochroleuco vel aureo; stromate aerugineo fere nigrescenti, chlamydosporis terminalibus, intercalaribus, singulis, binis, catenulatis vel acervalibus. Status conidicus Hypomycetum sect. Pseudomartiellae. 842 Journal of Agricultural Research vol. xxx, NO9

cc. Terminal chlamydospores absent. d. Intercalary chlamydospores present. e. Sporodochia typically absent. Conidia, when free- borne on aerial mycelium, spindle-shaped. Macroconidia gradually attenuate, generally lanceolate, apedicellate; but also sickleshaped and pedicellate; color intermediate between that of Roseum and Gibbosum sections, scle- rotia may be present Sec. 5, Arthrosporiella. ee. Sporodochia typically present. f. Macroconidia with top ends much attenu- ated; stroma typically brown,9 sometimes carmine10 Sec. 6, Gibbosum. ff. Macroconidia with top ends somewhat trun- cate; conidia ochreous to salmon; stroma Roseum-like; blue sclerotia may be pres- ent Sec. 10, Discolor. dd. Intercalary chlamydospores absent. e. Top ends of macroconidia gradually attenuate; when free-borne on aerial mycelium, sickle- shaped; or none. Acid-color modification of aerial mycelium yellow except in F. anthophilum (A. Br.) Wr., and other related fungi; conidial walls thin Sec. 7, Roseum. ee. Top ends of macroconidia somewhat constricted. f. Conidial walls thin,, and in this character, as well as in shape and color, similar to section Elegans Sec. 9, Lateritium. ff. Conidial walls thick, highly refracable, and in this character, as well as in shape and color, similar to section Discolor; conidial stage of Gibberella saubinetii and other similar Gibberellas Sec. 12, Saubinetii.

RELATIONSHIP OF FUSARIA TO ASCOMYCETES A number of different species of Fusaria have been definitely connected with certain ascomycetes of the Hypocreales group. In some other cases the connec- tion is very probable, and in still other cases the connection is suggested as possible. For convenience the instances are mentioned here under different sections of the genus Fusarium. Sec. 1. Eupionnotes. Perfect stage, Nectria moschata Glück, conidial stage * similar to F. aquaeductuum Lagh. var-. pusillum Wr. Sec. 2. Arachnites. Perfect stage, Galonectria graminicola (Berk. & Brome) Wr., conidial stage F. nivale (Fries) Ces. Sec. 3. Sporotrichiella. No connection with an ascomycete is known. Sec. 4. Camptospora. Perfect stage Nectria episphaeria (Tode) Fr., conidial stage may be F. cavispermum Cda. Sec. 5. Arthrosporiella. No definite connection with an ascomycete is known . Sec. 6. Gibbosum. No connection with an ascomycete is known. Sec. 7. Roseum. Gibberella tropicalis Rehm. possibly has Roseum-like conidial stage. Sec. 8. Liseola. Gibberella acervalis (Moug.) Wr., conidial stage very similar to F. moniliforme Sheld. Sec. 9. Lateritium. Gibberella baccata (Wallr.) Sacc, conidial stage F. later- itium Nees; G. pulicaris (FT.) Sacc, conidial stage F. sarcochroum (Desm.) Sacc; G. moricola (Ces. et Not.) Sacc, conidial stage F. urticearum (Cda.) Sacc; G. effusa Rehm, conidial stage F. salicis Fuck.; G. evonymi (Fuck.) Sacc, conidial stage F. pyrochroum (Desm.) Sacc; G. juniperi (Desm.) Wr., conidial stage F. fructigenum Fr. Sec. 10. Discolor. Gibberella heterochroma Wr., conidial stage F. polymorphum- like; and G. cyanogena (Desm.) Sacc, conidial stage F. sambucinum Fuck. Sec 11. Spicarioldes. No connection with an ascomycete is known. Sec 12. Saubinetii. Gibberella saubinetii (Mont.) Sacc, conidial stage F. çraminemum Schwabe; G. flacca (Wallr.) Sacc, ccnidial stage F. cariéis Oud.

9 Subsection Eugibbosum n. subsect. 10Subsect. Ferruginosum (Sect. Ferruginosum Sherb.) n. comb. May 1, 1925 Taxonomic Studies of Fusarium 843

Sec. 13. Elegans. No connection with an ascomycete is known. Sec. 14. Martiella. Hypomyces ipo?noeae (Hals.) Wr., conidial stage F. javanicam Koord.; Hypomyces cancri Wr., conidial stage F. striatum-like; Hypomyces leptosphaeriae (Niessl) Wr., conidial stage F. sphaeriae Fuck. Sec. 15. Ventricosum. Hypomyces solani Rke. et Berth., conidial stage F. argillaceum (Fr.) Sacc.

NOMENCLATURE IN RELATION TO LITERATURE CITED SPECIES, VARIETIES, AND FORMS (1) APPEL, O., and WOLLENWEBER, H. W. 1910. GRUNDLAGEN EINER MONOGRAPHIE DER The plan recently considered and GATTUNG FUSARIUM (LINK). Art). K. Biol. approved by the committees of the Anst. Land-u. Forstw. 8: 1-207, illus. Phytopathological Society, the Society (2) and WOLLENWEBER, H. W. 1910. DIE KULTUR ALS GRUNDLAGE ZUR BESSEREN of Agronomy, and the Mycological UNTERSCHEIDUNG SYSTEMATISCH SCHWIERIGER section of the Botanical Society on HYPHOMYCETEN. Ber. Deut. Bot. Gesell. June 6, 1924, in Washington, D. C, 28: 435-448, illus. (3) BROWN, W., and HöRNE, A. S. is followed in regard to questions of 1924. STUDIES OF THE GENUS FUSARIUM. I. nomenclature and terminology. This GENERAL ACCOUNT. Ann. Bot. 38: 379-383. plan provides for the use of the Latin (4) JOHANN, II. trinomial composed of the genus, 1923. INFLUENCE OF TExMPERATURE ON THE MORPHOLOGY OF FUSARIUM SPORES. (Abstract) species, and variety. The species in- Phytopathology 13: 51. cludes groups of individuals which can (5) SHERBAKOFF, C. D. be separated on the basis of morpho- 1915. FUSARiA OF POTATOES. N. Y. Cornell logical character of such a nature as Agr. Exp. Sta. Mem. 6: 97-270, illus. to be applicable and usable by mycol- (6) 1922. FUSARIA OF WHEAT AND CORN. (Abstract) ogists in general and which will be Phytopathology 12:45. most serviceable for practical purposes. (7) • ■ The variety is distinguished by morpho- 1925. USE OF AGAR BLOCKS FOR PHOTOGRAPHING logical characters, but less important LIVING SPORES. (Abstract) Phytopathology than those used for specific segregation. 15:44. An additional category termed forma (8) WOLLENWEBER, H. W. 1913. STUDIES ON THE FUSARIUM PROBLEM. is to be applied to subdivisions of the Phytopathology 3: 24-50, illus. species or varieties characterized and (9) distinguished primarily by physio- 1914. IDENTIFICATION OF SPECIES OF FUSARIUM OCCURRING ON THE SWEET POTATO, IPOMOEA logical instead of morphological char- BATATAS. Jour. Agr. Research 2: 251-286, acters, though in some instances there illus. may be present also some slight mor- (10) 1917. FUSARIA AUTOGRAPHICE DELINEATA. phological, distinguishable differences. Ann. M y col. 15: 1-56. The forma is designated by an arabic (11) numeral. 1924. PYRENOMYCETEN-STUDIEN. Angew. Bot. Ztschr. Erforsch. Nutzpflanz, 300-313, illus.