CHAPTER1 Nomenclature, and Identification Roy Watling, Glen A. Kile, and Harold H. Burdsall, Jr,

onfusion has surrounded the nomenclature groups, termed “biological ”, could be recog- and taxonomy of the Armillaria nized in Europe (Korhonen 1978, 1980) although, as (Fr.:Fr.) Staude for over a century. Until re- such, “biological species” had no standing within the cently, taxonomists have consistently dis- International Code of Botanical Nomenclature. Ander- agreedC on the exact description of the genus and its son and Ullrich (1979) expanded this approach using correct name according to the International Code of Bo- North American isolates. Morphological and genetic tanical Nomenclature. This confusion has created un- data have subsequently been combined to link many certainty for taxonomists and pathologists, and biological species to morphological species and vice has hindered the study of this widely distributed and versa (see chapter 2). Many laboratories now consis- economically important genus of fungi. Based on the tently test interfertility to identify unknown vegetative analyses of Watling and others (1982), we consider the isolates. genus to be a natural grouping, and that Armillaria is the appropriate generic name. This conclusion has been Armillaria probably contains about 40 species, of which widely accepted since that publication (Antonin 1986, several may have restricted geographical distributions Bérubé and Dessureault 1988, Guillaumin and others or vegetation associations. The movement of phanerog- 1985, Intini 1988, Marziano and others 1987, Rishbeth ams or their products from one area of the world to an- 1983, Roll-Hansen 1985, Romagnesi and Marxmuller other may, however, have extended distributions of 1983, Termorshuizen and Arnolds 1987). some species.

The first record of an Armillaria species was probably Species of Armillaria are necrotrophic pathogens of either in 1729 (Micheli) or 1755 (Battarra). However, , and in one case of another agaric, and not until the later classical authors began to describe mycotrophic associates of achlorophyllous plants (see the larger fungi could several taxa now assigned to chapter 8). Some ecological niches recorded for mem- Armillaria in its restricted sense (Armillaria sensu stricto) bers of the genus are undoubtedly exploited by several be unequivocally recognized. From the pathologists’ species, but the formerly broad concept of A. mellea ap- viewpoint, confusion has arisen from the assumption plied by many authors has confounded recognition of of many authors that Armillaria mellea (Vahl: Fr.) the species involved. Retaining voucher specimens of Kummer is a single variable or polymorphic species basidiomes 1 and vegetative isolates from phytopatho- (Singer 1956) that occurs in both temperate and tropical logical studies is thus extremely important. Although regions. Although this contention is supported by the ability to identify species of Armillaria has ad- maps purporting to show worldwide distribution (Dis- vanced rapidly only in recent years, we have accumu- tribution of Plant Diseases 143, 3rd ed. 1969) and by lated a wealth of observational and experimental host lists on an international or local basis (Laemmlen information on various aspects of Armillaria biology and Bega 1974, Pegler and Gibson 1972, Raabe 1962a), which makes it one of our better-known genera of classical European authors such as Bolton (1788-91) re- . alized that several taxa were involved. Nomenclatural and taxonomic aspects of Armillaria in European interest in morphological studies of general and the European species in particular have Armillaria was renewed in the 1970’s (Romagnesi 1970, been amply described in recent years (Antonin 1986, 1973, 1978; Singer 1970a,b; Singer and Clemencon 1972). The demonstration of a bifactorial sexual incom- 1The term basidiome is used in this publication in preference to less patibility system in an Armillaria species (Hintikka specific terms such as basidiocarp, carpophore, fructification, 1973) led to studies that showed several intersterile body, fruiting body, sporocarp, sporophore (Maas Geesteranus 1971),

Taxonomy and Identification 1 Guillaumin and others 1985, Herink 1973, Marxmüller The last two are now considered species of Tricholoma 1987, Roll-Hansen 1985, Romagnesi and Marxmüller (Fr.) Staude, and Ag. mucidus is placed in Oudemansiella 1983, Termorshuizen and Arnolds 1987, Watling 1987, Spegazzini (or Mucidula Pat.). Adopting either Ag. Watling and others 1982). This chapter provides an in- aurantius or Ag. robustus as the type could lead to troductory survey of the major issues in the nomencla- Armillaria becoming a synonym of Tricholoma. Kuhner ture and taxonomy of the genus. (1988) suggested Ag. mucidus as the type, but this was never recommended by any earlier author. This choice would be unfortunate as Ag. mucidus has little in com- Armillaria (Fr.:Fr.) Staude— mon with Ag. melleus. The selection of Ag. melleus as Nomenclature and Typification type by Clements and Shear (1931), Dennis and others (1954), and Donk (1949, 1962) was supported by In Fries’ Systema Mycologicum (1821), 12 species, includ- Watling and others (1982). Agaricus melleus Vahl:Fr. is ing Agaricus melleus, were accepted in the tribe based on Icones plantarum, Flora Danica (1792), vol. Armillaria, which he had established 2 years earlier 6(17): 9, plate 1013 (1790), M. Vahl (fig. 1.1) [= Armil- (Fries 1819). The tribes Armillaria and Lepiota were later laria mellea (Vahl:Fr.) Kummer in Der Fuhrer in die combined (Fries 1825) with the latter name used for the Pilzkunde (1871)]. As no herbarium specimen was avail- enlarged group. However, Fries (1838) reverted to able to support this plate, neotypic material was desig- Armillaria for some species. By this time, the number of nated (Watling and others 1982). species in the tribe had doubled, but its scope remained unchanged in his later Monographia Armillariarum The generic name Armillariella (Karsten 1881) typified Suecicae (Fries 1854). by Ag. melleus has been used in many publications; if Armillaria is based on a species other than Ag. melleus, Staude (1857) was the first to raise Fries’ tribe to ge- Armillariella would become available. Karsten’s genus neric rank. Singer (1951b, 1955a,b, 1986) has disputed is logical if Armillaria is typified by Ag. luteovirens Alb. whether Staude’s entry meets all the requirements for valid publication, but Staude is now generally accepted as the validating author of the genus (Donk 1949, 1962; Watling and others 1982). Singer (1951b, 1955a,b) pro- posed Kummer (1871) as the correct author for Armillaria, and has recently reiterated that belief (Singer 1986), a conclusion we do not accept. Thus, Singer (1986) has argued that the publication of Staude (1857) is inadmissible according to the International Code of Botanical Nomenclature, but nothing has changed since Donk (1949, 1962) clearly discussed the status of Staude’s account. Watling and others (1982) found no reason to disagree with Donks findings. Both Staude and Kummer (1871) include within their ge- neric concept Agaricus melleus, and as far as anyone can decide from the available information, it agrees with that outlined within Fries’ (1821) tribe Armillaria. Fries (1821; p. 26) includes a reference to Battara (1755) un- der synonymy of tribe Armillaria but nowhere dis- cusses this entry further. We think that this one mention can hardly support Singer’s statement “de- fines the basic scope of the tribus.” Nothing in Fries (1821) or in Battara (1755) necessitates further explora- tion, and this re-emphasizes the importance of Systems Mycologicum (Fries 1821) in forming a clear base line. Clements and Shear (1931) subsequently selected it as type species for the genus in their comprehensive sur- vey of the nomenclature of the genera of fungi.

After accepting Staude’s authority for Armillaria, the FIGURE 1.1 — Agaricus melleus, as illustrated by Martin Vahl in typification of the genus follows in a straightforward Flora Danica (1790 - 1792). Marxmuller and Printz (1982) manner. Staude (1857) included four species: Ag. considered this figure could also represent Armillaria borealis, mucidus, Ag. melleus, Ag. aurantius, and Ag. robustus. although Marxmuller (1987) accepted it as Agaricus melleus.

2 Taxonomy and Identification & Schw.:Fr., as supported by Singer (1951a). However, expallant, hygrophanous or not; color variable yel- this species was not originally in Fries’ tribe, a prereq- low-brown, yellow-olivaceous, ochraceous, rusty- uisite for consideration. Armillariella is therefore an ob- tawny, umber, cigar brown, less commonly buff or ligate synonym of Armillaria. Floccularia Pouzar is the clay pink, sometimes ivory, pallid, or even mouse- correct genus for Ag. luteovirens and its allies. gray; surface glabrous, scurfy, squamulose, squamules darker than ground color, sometimes re- Incorporating Armillaria into Clitocybe (Fr.) Staude has stricted to disc; glabrescent as scales are lost; dry or sometimes blurred the identity of what we believe to becoming viscid to distinctly viscid, in some species be a natural genus. While first proposed by Ricken almost glutinous. (1915), French mycologists have most frequently fol- Stipe — central, fibrous-fleshy, not characteristically lowed this approach, for example Kühner and cartilaginous; often becoming hollow and the outer- Romagnesi (1953) and Heim (1950, 1963). The latter in- most layers splitting and curling back to expose cluded tropical species of Armillaria in his concept of flesh; more or less annulate with floccose-membra- Clitocybe. This proposal does not interfere with typifica- nous to arachnoid veil; often arising from sheets of tion, as Armillaria would simply become a synonym of white mycelia or from well-differentiated black Clitocybe. However, clear differences exist in basidiome rhizomorphs, and/or, associated with plaques of development between Armillaria and Clitocybe (Watling thin, black, tough tissue. and others 1982). Additionally, Bennell and others Lamellae — close to subdistant; moderately thick; (1985) showed radical differences in basidiospore wall nearly white, ivory, or cream-color at first but fre- morphology between A. mellea and Clitocybe nebularis quently becoming spotted with cinnamon-buff, (Batsch:Fr.) Kummer. Clitocybe tabescens (Scop.:Fr.) rusty-tawny, or sometimes, particularly with age, Bres. is the species usually cited as a link between the with a tinge of purple or distinctly pink; sinuate; two genera. It is similar to A. mellea in basidiome devel- adnexed to deeply decurrent. opment, basidiospore wall structure, and its bifactorial Flesh — of pileus pale and of stipe white at first, be- heterothallism (Anderson 1982). This species is thus coming as dark as umber or Vandyke brown down- best placed in Armillaria, probably as A. socialis (DC:Fr.) wards and sometimes tinted red or bluish at base Herink [synonym A. tabescens (Scop.:Fr.) Emel.]. where colonized by pigment-producing bacteria or nectriaceous fungi. Singer (1951a) and Herink (1973) suggested subcatego- -print — white to cream-color darkening ries of the genus. Singer divided Armillaria (as slightly on drying, and in herbarium material. Armillariella) into two sections distinguished by the Basidia — 4-spored, sometimes 2-spored; thin- presence or absence of a veil (annulate and exannulate walled; with or without a basal clamp-connection; species), a subdivision he later maintained (Singer hyaline; smooth-walled in aqueous alkali solutions 1986). Herink (1973) followed Singer and recognized or if thick-walled [= crassobasidia (Chandra and Armillaria as an annulate subgenus and Desarmillaria as Watling 1983)] then appearing silvery or glassy, an exannulate subgenus. He placed Armillaria mellea in and/or, becoming ochraceous or fulvous. the first and A. socialis in the second. His ideas agree Basidiospores — ellipsoid; inamyloid; hyaline, yel- with our own concepts, although we believe it will lowish cream-color or ochraceous in aqueous alkali eventually be possible to subdivide the subgenus solutions; weakly cyanophilic; thin to moderately Armillaria into related subgroups. thick-walled; smooth or slightly verruculose or rugulose with broad, blunt usually prominent apiculus; lacking germ-pore or apical differentiation Generic Characteristics (thinning or thickening). Cheilocystidia — present or absent, often incon- Various morphological, cultural, and other features spicuous; variable in shape sometimes catenulate- help distinguish Armillaria from other genera of septate; thin-walled or becoming slightly Agaricales. Collectively, these characters define the ge- thick-walled with age sometimes with apical prolon- nus, and variations among them define species. The gation and with or without basal clamp-connection; following are the salient characteristics of Armillaria: smooth; hyaline to honey-colored in aqueous alkali solutions. Habit — clitocyboid with slightly sinuate, adnexed, Pleurocystidia — absent or, if present, thin-walled; subdecurrent or decurrent gills; bivelangiocarpic or poorly differentiated and rarely visible above the metavelangiocarpic development in annulate spe- level of the basidia. cies, apparently monovelangiocarpic development Pileipellis — an irregular, disrupted trichodermium in exannulate species; solitary, gregarious, or cae- consisting of (i) an irregular, easily destroyed spitose. suprapellis composed of groups of fulvous or cinna- Pileus — fleshy, thinning towards margin, mon, subparallel, ascendant, loosely to strongly

Taxonomy and Identification 3 adhering hyphae intermixed with broad, frequently The many distinctive morphological characteristics of encrusted hyphae (which form the scales), often the genus, the production of characteristic rhizo- with clamp-connections; ascendant hyphae becom- morphs, both parasitic and saprophytic capabilities, ing repent to form a rather amorphous adnate layer; and the apparently diploid nuclei in the vegetative my- (ii) mediopellis - of parallel to subparallel hyphae celium (see chapter 2), lead us to believe that it forming a cutis that mayor may not gelatinize but stands quite distantly from other agaricoid genera. sooner or later becomes the outermost layer; and (iii) Thus, Jülich’s (1981) introduction of the family Armil- subpellis - a compact hyphal layer. lariaceae to accommodate the genus has great merit. Stipitipellis — parallel hyphae overlain by more or less strongly developed, irregular, filamentous velar remnants; in parts of stipe free from velar material Relationships Within Armillaria showing development of cylindric to elongate clav- ate or lageniform caulocystidia. Apart from the subgeneric distinction between devel- Pileus and stipe trama — monomitic; hyphae opmental patterns in annulate and exannulate species inamyloid, generally lacking clamp connections. and its inference of relatedness, no systematic attempt Hymenophoral trama — bilateral at first and re- has been made to assess the phylogeny of species based maining so or becoming regular with age although on differences in morphology, physiology, biochemis- always demonstrating some divergent arrangement; try, ecology, pathology, or sexual compatibility system. constitutive hyphae generally lacking clamp-connec- Computer-aided comparative studies of such attributes tions; inamyloid. could assist research into species relatedness. Vegetative growth — variable on agar media but typically reddish-brown crustose surface mycelium; Divergent nucleic acid composition has probable utility usually slow growing; with or without tufts of cin- in ascertaining species relatedness. Anderson and oth- namon aerial mycelium; with or without reddish- ers (1987) concluded that some particular DNA se- brown rhizomorphs or with white to cream-color quences may be appropriately variable for rhizomorphs embedded in the medium with phylogenetic studies. Subsequently, Anderson and oth- emergent reddish-brown tips; rhizomorphs branch ers (1989) showed that some European Armillaria spe- monopodially, dichotomously, or irregularly; veg- cies and the equivalent or unidentified North American etative mycelium often bioluminescent; cells uni- or Biological Species, or NABS, (Anderson and Ullrich multinucleate; nuclei apparently diploid. 1979; Bérubé and Dessureault 1988, 1989) could be Rhizomorphs — mycelial aggregations with a mela- placed in distinct classes based on restriction maps of nized outer layer and pale, apical growing tip; pro- ribosomal DNA. These are: rDNA class 1, A. ostoyae duced in culture and from infected lignicolous (= NABS I); class 2, A. gemina (= NABS II); class 3, A. bo- material. realis; class 4, A. sinapina (= NABS V); NABS IX, X; class Single basidiospore isolates — from heterothallic 5, A. calvescens (= NABS III), A. gallica (= NABS VII), A. species typically slow growing; producing white, cepistipes (= NABS XI?); class 6, A. mellea (= NABS VI). fluffy to cottony mycelium, sometimes with areas The classes are believed, with the possible exception of of brown or reddish; with or without sparse rDNA class 4, to represent natural groupings. In addi- rhizomorph development; nuclei haploid. tion, classes 1,2, and 3 were considered to be closely Compatibility system — bifactorial; heterothallic related with rDNA classes 2 and 3 derived from the with multiple alleles at the incompatibility loci; more widely distributed DNA class 1. Greater resolu- some species possibly homothallic. tion through detailed mapping of particular regions of the genome will assist phylogeny development. As Anderson and others (1989) have suggested, reconsid- Relationships With Other Agarics ering ecological, morphological, and distributional data for taxa on the basis of restriction polymorphisms Modern classifications of the Agaricales link Armillaria would be informative. S.S. with the Tricholomataceae (Jülich 1981; Kühner 1980; Singer 1951a, 1986). However, even in the tem- perate northern hemisphere where the agarics have Present and Excluded Species of been most intensively studied, only Jülich (1981) indi- Armillaria cated a strong relationship between Armillaria and an- other genus in the Tricholomataceae, Tricholvmopsis Singer (1978) prepared a key to the world taxa (as Singer. Possible relationships to the Cystodermataceae Armillariella) he considered distinct. This key needs to (Romagnesi 1980), the Entolomateaceae (Bennell and be updated in light of the new taxa recognized and others 1985), and the (Helfer and Watling concepts developed since that time. Table 1.1 lists 36 1989) also have been discussed. taxa which we believe have been documented suffi-

4 Taxonomy and Identification Taxonomy and Identification 5 ciently to be considered species, although a few addi- Recent major contributions to the description of mor- tional taxa will probably be delineated eventually. It in- phological variation and the delineation of Armillaria cludes all species known to be significant to plant taxa include those of Romagnesi (1970, 1973, 1978); pathologists and ecologists. Nomenclatural adjustment Marxmüller (1982, 1987); Marxmüller and Printz (1982); of some of Singer’s Armillariella species is required. Romagnesi and Marxmüller (1983); Watling (1987) for Fourteen of the species have been recognized as both Europe; Singer (1956, 1969) for South America; morphological and biological species (see chapter 2), Stevenson (1964) and Kile and Watling (1981, 1983, and future interfertility-morphological studies may re- 1988) for Australasia; and Bérubé and Dessureault sult in changes to the status of other species listed in (1988, 1989) for North America. Although Chandra and table 1.1. Watling (1982) redescribed several Indian species, fresh collections are required to complement their herbarium Since Fries (1821), many species have been placed in studies. Mohammed and others (1989) and Mwangi Armillaria by virtue of possessing a white to cream- and others (1989) reported cultural, genetic, and color spore-print and an annulus, which make it very isozyme studies of African species which will help to heterogenous. With a more restricted generic concept resolve their identity. Further research is necessary for for Armillaria, knowing where some of these taxa for- other areas such as Siberia, China, and parts of South- merly placed within Armillaria are now assigned is use- east Asia. ful. Table 1.2 shows the concordance of Fries (1821) species with modern concepts. Fries (1838, 1854, 1874) included an additional 34 species in the Armillaria Taxonomic Characters and Identification group, only one of which was possibly an Armillaria species S.S. (A. laricinus = A. ostoyae?). Many velate spe- As with other macromycetes, species of Armillaria are cies of Tricholoma have been placed in Armillaria, and T. delimited primarily by basidiome morphology (fig. caligata (Viv.) Rick. and its allies have been traditionally 1.2). While vegetative isolates may be identified or placed by North Americans in the genus (Hotson 1941, grouped by various methods, basidiomes are essential Mitchel and Smith 1976, Smith 1979, Thiers and for the complete description and naming of species. Sundberg 1976). This is erroneous and confusing be- cause the species are morphologically, ecologically, and Basidiome macromorphology, pileipellis structure and biologically quite distinct from Armillaria species S.S. ornamentation, ring characteristics, stipe ornamenta- tion, presence or absence of subhymenial or basidial Romagnesi (1970, 1973), Termorshuizen and Arnolds clamps, location of pigments in cell walls or vacuoles, (1987), Watling (1987), and Watling and others (1982) and basidiospore size and ornamentation are among discussed the identity of Armillaria species illustrated characters of value for species differentiation. Separa- in the classical literature. tion of some species by morphological criteria alone is difficult but no more so than in many other agaric gen- era. Identification may require using numerous macro- and micromorphological features combined with bio- chemical, cultural, and ecological information. A thor- ough appreciation of the most useful taxonomic characters will only be derived from careful analyses of all these features (Watling and others 1982).

Analysis of European, and to a lesser extent Australasian species (Kile and Watling 1983, Shaw and others 1981), showed that it is possible to identify some species by morphological and physiological attributes of their vegetative mycelia and rhizomorphs as well as by basidiome morphology (table 1.3). Additional simple tests such as the response of the mycelium to light may also differentiate some species (Benjamin 1983; see also Hood and Sandberg 1987).

Serological differences among several Armillaria species were demonstrated by Lung-Escarmant and others (1978, 1985b) and Lung-Escarmant and Dunez (1979, 1980); serological techniques may, in the future, have a substantial impact on the delimitation of Armillaria

6 Taxonomy and Identification A B C

D E F

G H I

J K

L

Taxonomy and Identification 7 species. Fox and Hahne (1989) used monoclinal anti- gallica. Jahnke and others (1987) and Anderson and oth- bodies, but the results to date are not as impressive as ers (1987, 1989) showed that mitochondrial (mt) DNA those obtained by studies using polyclonal antibodies. was highly conserved within species but divergent be- Refinement of the techniques by developing greater an- tween them, and that restriction fragment patterns tibody specificity to overcome problems of cross reac- were therefore diagnostic for species. Smith and tivity between closely related species may allow Anderson (1989) correctly identified 23 North Ameri- accurate identification in the near future, including the can isolates using DNA restriction fragment length possibility of diagnostic kits for rapidly identifying polymorphisms. field material. Isoenzyme and protein profiles of some northern hemi- Nucleic acid analysis supports current species concepts sphere taxa also differ sufficiently to offer further in Armillaria, and offers a powerful diagnostic tool. methods of species separation (Lin and others 1989, Motta and others (1986) reported quantitative differ- Lung-Escarmant and others 1985b, Morrison and oth- ences in nuclear DNA content between A. mellea and A. ers 1984).

8 Taxonomy and identification The biological species concept has been applied to the enhanced their individual values; the frequent concor- genus using single basidiospore isolates to delineate re- dance of information from a variety of sources has more productively isolated groups as discussed in chapter 2. clearly defined many taxa. Additional collections and Using this particular approach has greatly assisted tax- application of various techniques to assess phenotypic onomists in defining species in genera with restricted and genotypic variation within the Armillaria flora in re- interspecific but high intraspecific morphological varia- gions where it is incompletely known remain necessary tion. Reproductively isolated groups have been linked to enhance our taxonomic understanding of the genus to existing taxa (Marxmüller 1982, Romagnesi and on a worldwide basis. Analysis of collections on which Marxmüller 1983), led to the description of new taxa some early names are based will further assist the quest (Bérubé and Dessureault 1988, 1989; Marxmüller and for nomenclatural stability within the genus. Korhonen in Marxmüller 1982; Marxmüller 1987), and established intra- (Anderson and Ullrich 1979, Kile and The genetic approach to species differentiation, initiated others 1983, Korhonen 1978) and inter-continental dis- for Armillaria by Korhonen (1978), allowed the identifi- tributions (Anderson and others 1980, Guillaumin and cation of species from vegetative isolates. Subsequent others 1989a, Morrison and others 1985a). Conversely, work has shown that vegetative isolates also may be species initially described on conventional criteria were distinguished by other cultural or physiological charac- later shown to be biological species (Guillaumin 1986a, teristics. The ability to identify vegetative isolates is Kile and Watling 1988). highly useful for organisms in which the vegetative phase may often be the only one encountered. Newer Cumulative experience suggests that reconciling mor- techniques such as DNA analysis and production of phological (taxonomic) and biological species concepts monoclinal antibodies have the potential to further en- for most Armillariae will be possible. Although such hance rapid and reliable identification of vegetative iso- studies will take time to complete, they should result in lates. robust characterization of species. In cases for which detailed morphological examination supports a single The morphological and biological species concepts ap- species but interfertility studies indicate otherwise, pear largely reconcilable for Armillaria, at least on the Watling and others (1982) suggested adopting the basis of our knowledge of temperate species. This per- macro-microspecies concept in which a macrospecies haps fortuitous situation will continue to have a marked would consist of morphologically indistinguishable impact in clarifying the taxonomy of the genus. biological species. We support this suggestion. A stable nomenclature, well-defined species, and a vari- ety of identification techniques are invaluable to pa- Conclusions thologists and ecologists in their attempts to understand the behavior and natural relationships of Armillaria spe- Major studies of Armillaria taxonomy have been com- cies, clarify their natural relationships, and develop dis- pleted in recent years. Linking morphological, cul- ease-control strategies. Progress has been significant in tural, physiological, and genetic data has often the former areas in recent years.

Taxonomy and Identification 9