THE GENUS FLAMMULINA
A TENNESSEE TUTORIAL
Ronald H. Petersen1, Karen W. Hughes1, and Scott A. Redhead2
1Botany Department, University of Tennessee,
Knoxville, TN 37996-1100. repete@utk.edu, khughes@utk.edu
2 Mycological
Herbarium,, Research Center, Central Experimental Farm, Ottawa, K1A
0C6, Canada. redheads@em.agr.ca
|
|
THE GENUS FLAMMULINA - INTRODUCTION
Karsten. 1891. Meddel. Soc. Fauna & Flora Fennica 18: 62.
For the casual mycophagist, Flammulina has been deceptively easy. The "fuzzy foot" mushroom (a.k.a "velvet shank") has been a staple in the diet, often available in the winter months when most mushrooms take a rest. Moreover, in recent years an abnormal growth form has become a supermarket product called enoke-take, following the Japanese origin of the product. For the mushroom taxonomist, however, the plot has thickened considerably. It is no longer adequate to assume that all collections are F. velutipes, and whats more, the edibility of some species is untested or folkloristic at best.
A long and involved study of the genus was undertaken at the Tennessee mycology lab, attempting to make some sense of the genus. To be sure, our study was preceded by proposal of two species (F. fennae, F. ononidis; see discussion below), and a number of mating studies (Lamoure, 1989; Yokoyama, 1992) which left murky results. Physiological experiments using cultures of putative Flammulina taxa (Klan et al., 1992; Psurtseva and Mnoukhina, 1998) seemed conclusive but still left unanswered questions. A very preliminary experiment in our lab using isolates from widely scattered locations in the North Temperate Zone confirmed that at least two intersterility groups within "F. velutipes" could be identified. Casual conversations with Scott Redhead revealed that he had studied hundreds of herbarium specimens and he also suspected that additional taxa lurked among them.
The race was on. In this day, such an investigation must be complemented by molecular studies (read DNA), a powerful tool which can segregate species and populations and reveal much about the evolution of the group under scrutiny. So a team was assembled: Scott Redhead on morphology, Ron Petersen on compatibility tests, Karen Hughes and Andy Methven on molecular considerations, and Nadya Psurtseva on physiology.
Once the study took shape, we broadcasted an appeal for materials from both the professional (i.e. personal contacts, cultures collections, etc.) and amateur communities (i.e. in club newsletters, MycoInfo, etc.). The result was gratifying: cultures, dried and fresh basidiomata, and spore prints arrived, so the eventual study incorporated over 200 strains worldwide.
In many cases, some item was missing: dried specimens without spore print, culture or spore print without fruitbodies, etc. Andy Methven (Eastern Illinois University) and Nadya Psurtseva (Komarov Institute, St. Petersburg, Russia) volunteered to fruit the cultures in our hands, and did so on tulip poplar (Liriodendron tulipifera) sawdust supplemented with rice bran. The results can be seen in photographs below.
Now, unfortunately, this section will become a bit complicated, for it is necessary to spell out the methods underlying mating and molecular studies. If this stuff is boring, skip to the next section, called "Taxonomic characters."
Mating studies. In order to perform mating experiments, it is necessary to use haploid cultures as the gametic parents. Luckily, there were two means to obtain such cultures: 1) from spores; or 2) from conidia.
The first method again took two forms: 1) from a fresh basidiome a small wedge of pileus (including a few lamellae) was suspended (using a small blob of Vaseline) under a Petri dish cover and spores were dropped on malt extract (Difco, 1.5%) agar (Difco bacto-, 2.0%; MEA). Once spores germinated, they could be isolated into separate cultures for use in mating tests. When cultures were fruited on sawdust, the resultant basidiomata could be used to establish single-spore isolates. 2) From spore prints, spores were diluted in sterile distilled water and spread on MEA. Excess water was decanted, and plates allowed to dry overnight to remove the remaining film of water. Once spores germinated, separate single-basidiospore isolates (SBIs) were harvested.
The second, fortuitous, trait of Flammulina is the production of conidia (arthrospores; asexual propagules) in cultures. An even more serendipitous trait is that dikaryon cultures (i.e. from fruitbody flesh, cultures collections, etc.) produce conidia in such a way that the conidia are monokaryotic (Aschen, 1952; Kemp, 1980; Ingold, 1980), even though the parent mycelium is dikaryotic. So single-conidial cultures could be obtained easily and used in mating experiments. In all, we had over 100 sets of haploid isolates with which to work, in most cases comprising 15-20 individual cultures.
For mating experiments, fully mature haploid isolates (often called, incorrectly, "monokaryon," without regard to the chromosomal number a monokaryon could be diploid) on MEA were disked with a sterile cork borer. Disks of each participant mycelium were placed 6-9 mm apart on MEA, and resultant mycelia were allowed to grow, make contact, and overgrow in the "contact zone." After some time, contact zones were inspected in a search for clamp connections, a reliable sign of sexual compatibility.
Several kinds of mating experiments were performed:
Molecular techniques. Not all species in Flammulina were available for DNA extraction. For example, F. callistosporioides and F. similis could not be included because both species are known only from their type collections and perhaps one auxiliary specimen. Species that were subjected to molecular experiments were F. velutipes, F. fennae (Horak isolates), F. mexicana, F. populicola, F. rossica, F. elastica, and F. ononidis. Care was taken to include the exemplars designated for mating experiments (Petersen et al., 1999).
ITS sequences and phylogenetic reconstruction:
RFLP signatures: A study was designed (Methven et al., 2000) to identify species-specific restriction patterns, and used enzymes HAE III and Bst F51, cleaving the ITS1-5.8s-ITS2 DNA sequence (Fig. X). Restriction with HAE III revealed three patterns: DNA of all collections was cleaved at least once. Pattern 1 produced two fragments, pattern 2 produced three fragments, and pattern 3 produced four fragments. These patterns alone were not species-specific. Bst F51 produced two patterns: not all DNA was cleaved, and the pattern in which DNA remained intact was pattern 1. Pattern 2 included one cleavage site (two fragments). Thus the following patterns were noted (with HAE III represented by the first number and Bst F51 by the second): 1-0, 1-1, 2-0, 2-1, 3-0, 3-1. Using these two enzymes in tandem produced virtually species-specific patterns.
Within the many collections of F. velutipes, two additional restriction enzymes were applied in order to identify possible geographically circumscribed populations. These enzymes were Bgl I and Bst UI. Using Bgl I, two patterns were found: uncleaved DNA was called pattern A, once-cleaved DNA (two fragments) was called pattern B. Restriction with Bst UI revealed that DNA from all collections included a single cleavage site: this was uninformative and was given no pattern name. Some collections, however, showed an additional cleavage site. The pattern with only the universal cleavage site was called C, while the pattern showing two cleavage sites was termed D. With some obvious ambiguity, for F. velutipes the combination of these patterns (A-D) could be correlated to geographic localities.
Likewise, RFLP patterns could identify continental populations of F. populicola, one from the northern Rocky Mountains and Scandinavia, the other from Alaska and New Mexico.
CIRCUMSCRIPTION OF THE XERULACEAE
Over the past decade, the Tennessee lab has investigated mating systems and patterns in several genera placed in the Xerulaceae by Redhead (1987): Xerula (Petersen and Methven, 1994), Oudemansiella (Petersen and Halling, 1993), Strobilurus (Petersen, 1995). This family name, however, was coined by Jόlich (1981) with Lampteromyces (now thought to belong to the Paxillaceae), Oudemansiella, and Xerula as "examples" (not necessarily barring other genera). In the opinion of some authors, Rhodotus should be taxonomically included with Jόlichs exemplary genera, in which case the family name Rhodotaceae Kόhner (1980) is prior. For Kόhner and Jόlich, the Rhodotaceae was monotypic. Still further, Physalacria (at least some species) might be included based on phylogenetic reconstructions (J. M. Montcalvo, pers. comm.). In this case, the taxonomically inclusive family name would be Physalacriaceae Corner (1970). Whether Hormomitaria or Pseudotyphula would also be part of this complex (as accepted by Corner) remains unknown.
Especially in the latter case, Jόlichs (1981) circumscription of the Xerulaceae would require significant emendation. It is worthwhile to repeat Jόlichs family circumscription:
"Basidiocarp annual, agaricoid, stipitate, lamellate, with collybioid habit, pale to brown, dry or viscid to glutinous, a ring present or absent. Lamellae somewhat thick, broad, with regular trama. Context white, inamyloid. Hyphal system monomitic, in some species ± sarcodimitic. Hyphae hyaline, cylindrical, very narrow to large, thin- to slightly thick-walled in the subhymenium, somewhat to distinctly thick-walled in the stipe, smooth, usually with clamps. Cystidia large, hyaline, broad, smooth, thin- to somewhat thick-walled, smooth, somewhat guttulate. Basidia large, hyaline, clavate, smooth, thin-walled, guttulate, with a basal clamp, (2-) 4-spored. Spore-print white. Spores large, hyaline to yellow, globose to broadly ellipsoid, smooth to rough or minutely warted, slightly to distinctly thick-walled, guttulate, inamyloid."
Jόlich. 1981: 70 (under Tricholomatales)
"Xerulaceae. Agaricoid family, shares a common origin with the Collybiaceae, differs in its large basidia and large, thick-walled spores. Is probably related to the secotioid family Gigaspermaceae."
Furthermore, Jόlichs (1981: 284) diagnosis of the Rhodotaceae should be consulted. His note (page 71) about placement of Rhodotus reads: "Isolated position, perhaps related to the Hydnodontales." Inclusion of Rhodotus as an element within an expanded Xerulaceae ss. Jόlich introduces some new characters: 1) "bilateral [lamellar] trama," now known from at least three species of Flammulina; 2) absence of hymenial cystidia (not encountered in any other genus of the Xerulaceae to my knowledge); 3) pinkish-ochraceous spore print (other genera characterized by white spore-print); and 4) gelatinized pileus context, now also known in Flammulina. Nonetheless, Rhodotus represents pileate-stipitate agarics, and thus does not drastically distort the original circumscription of the Xerulaceae.
Inclusion of Physalacria, however, extends the family from strictly agaricoid to reduced forms lacking pileus, lamellae, pileipellis structure, etc. While perhaps justified from a phylogenetic viewpoint, such inclusion would drastically distort the intent of the Xerulaceae ss. Jόlich and the Rhodotaceae.
Larger than the problem of finding the proper circumscription for this family, the general tension between morphological similarity versus molecular relatedness and its influence on ranking and naming processes is highlighted. Another such example can be seen when coralloid fungi (Ramaria) and cantharelloid forms (Gomphus, Clavariadelphus) find relatives in the stinkhorns (Mutinus, Clathrus, Pseudocolus) (Pine et al, 1999; Humpert et al, 2000). Whether such divergent forms should be sheltered under a single family name will be debated.
More recently, Redhead (1987) emphasized presence of sarcodimitic basidiome tissues (most often in the stipe medullary region) in his circumscription of an emended Xerulaceae. Large numbers of species were surveyed for this feature, and because Rhodotus (and Flammulina) lacked sarcodimitic tissues, they were not included in Redheads sense of the family. Conversely, 25 genera were included, adding considerably to Jόlichs examples. More recently, however, Redhead and Norvell (1993) found sarcodimitic hyphal construction in fungi well outside the Xerulaceae, and questioned this character as diagnostic. Likewise, tissues quite similar to classic sarcodimitic construction are encountered in Clavaria s. str. (Petersen, 1968, as Clavulinopsis), not closely related to the Xerulaceae (see Pine, et al., 1999).
Corner (1966) included Trogia as part of the cantharelloid fungi, later (Corner, 1991) gathering taxa from several scattered genera into Trogia based on presence of sarcodimitic tissue and little else. Xerula was not merged into Trogia in spite of sarcodimitic stipe tissues.
Still later (Corner, 1996) subacerose basidiolar shape was used to diagnose yet another large generic complex, but Xerula was not included in that group also. His exposition on Xerula longipes (Corner, 1996), however, was based on very limited collections (five collections, totaling six basidiomata) from Mt. Kinabalu, Borneo, even though this epithet was proposed to represent European basidiomata
Petersen (2000) proposed Rhizomarasmius to represent two species formerly placed in Marasmius, R. pyrrhocephalus and R. undatus. Neither species exhibited sarcodimitic tissues, but both were characterized by typical xerulaceous cystidia, appropriate pileipellis structure, and radicating stipe. Corners description of Bornean collections of Xerula "longipes" closely fits the circumscription of Rhizomarasmius.
New additions. To define the Xerulaceae exclusively on the presence of sarcodimitic tissues is "one- character taxonomy." There are several other morphological character fields which may be informative: presence of characteristic hymenial cystidia and pileipellis arrangement are two. Close relationships as depicted in molecular phylogenetic reconstructions provide additional strong evidence. Using these characters, Flammulina (see Petersen et al., 1999), at least some species of Pseudohiatula, and Rhizomarasmius (Petersen, 2000) may be included in a further emended Xerulaceae. Molecular evidence also supports inclusion of Physalacria and Rhodotus (Montcalvo, pers.comm.).
Typification of Flammulina. When Karsten (1891) proposed the new genus name Flammulina to represent Agaricus velutipes, he transferred that species epithet to the new genus. The epithet was not his, of course, but had been proposed almost a century prior by William Curtis. Karstens action, therefore, was purely nomenclatural, not taxonomic.
In our study we came to recognize that Karstens herbarium included only F. elastica, giving him a distorted impression of the identity of F. velutipes. This misidentification of Finnish collections as F. velutipes was surely understandable, and had no effect on the nomenclatural process of proposing the new genus name. Curtiss A. velutipes (not Karstens impression of it) was the type species of Karstens genus.
In his study of type specimens of all generic names in the Agaricales, Horak (1968) examined specimens of what he thought was F. velutipes from Karstens herbarium. He was, of course, examining material of F. elastica, just as had Karsten. The necessary action was to examine English specimens from Curtis collections (non-existentl) or to designate a neotype for F. velutipes.
Curtiss proposal of Agaricus velutipes was accompanied by a handsome colored plate illustrating the species. Recognizing that the plate could be designated as lectotype (i.e. material in the hands of the original author when the name was proposed; in this case, an iconolectotype), Redhead & Petersen (1999) designated an epitype specimen including basidiomata (housed at Kew), spore print, and derived dikaryon and monokaryon cultures. Thus Flammulina now seems to rest on a type specimen with the possibility of further research on it.
TAXONOMIC CHARACTERS IN FLAMMULINA
A partial outline of some important taxonomic characters was furnished by Bas (1983) and Arnolds (1977) in papers proposing F. fennae and F. ononidis, respectively. After all, the new taxa had to be separated from F. velutipes, the commonly accepted species. Redhead, however, found new and unexplored characters which seemed useful for taxonomic separation, especially when correlated with beta-level results (i.e. mating studies; molecular experiments).
Pileipellis structure. Everyone knew that Flammulina was characterized by a slimy pileus surface, so the universality of this character was not surprising. The pileus surface layer (pileipellis) within and under the slime, however, seemed rather distinctive. In one group of species, the pileus surface layer was composed of sphaeropedunculate cells (i.e. the shape of a hot-air balloon) in a layer, together with longer, stiff, awl-shaped cells (i.e. pileicystidia). In the other group, the pileipellis comprised long, narrow, flexuous, often branched elements combined with pileicystidia. These tissue differences, when combined with stratose pileus trama and spore sizes, could be used to separate several new taxa (Redhead & Petersen, 1999) and taxonomically dispose of the rest of the genus.
Terminology, however, has been confusing, for various authors have used their own terms. A short summary is worth offering:
cuticle; pellicle; epicutis (all used by Singer); suprapellis (Bas) = the outermost layer of the pileus
hymeniiderm pileipellis = a more or less coherent layer of inflated hyphal tips forming the pileus surface (i.e.F. fennae
(ixo-)trichoderm = a turf of filamentous hyphal tips arising as such (i.e. F. velutipes) or from clavate bases as apical branches (i.e. F. rossica; F. populicola); ixo- = involved in slime
hypodermium (Singer); mediopellis (Bas) = in a two-layered pileus surface, the layer below the pileipellis, always gelatinized in Flammulina
subcutis (Singer); subpellis (Bas) = in a three-layered pileus surface, the layer between the epicutis and hypodermium
subhypodermium (Singer) = a tissue layer sometimes distinguishable between a hypodermium and pileus context
(ixo-)hyphidium; palisadenelement = the filamentous hyphal tips comprising the pileus surface (i.e. F. velutipes); ixo = involved in slime
pileicystidium; dermatocystidium; "hairs" = elongate-fusiform to cylindrical-lanceolate elements of the pileipellis, from widely to densely scattered, often varying with basidiome age
Stratose pileus trama. In North Temperate species of Flammulina, the pileus trama is more or less homogeneous from just under the thin hypodermium to the connection with lamellae. But a collection from New Zealand exhibited pileus trama tissue in which a layer of glassy, firmly gelatinous tissue was sandwiched between the pileus surface and lower pileus flesh. Such an arrangement is called "stratose" (i.e. layered, as in strata of rock), and the New Zealand species was proposed as new (F. stratosa; Redhead, et al., 1999). As the study progressed, additional specimens were found to show stratose pileus flesh. Two were from Argentina; F. callistosporioides and F. similis. Together with a single unnamed specimen from the Sierra Madre Oriental of Mexico, the complex exhibiting stratose pileus flesh appears relatively clearcut. The character has value, perhaps at subgeneric rank.
Lamellar trama structure. Up to the time of our study, known species of Flammulina all exhibited interwoven lamellar trama (i.e. the hyphae of the gill flesh were merely interwoven, without a particular orientation). Some species, notably those with stratose pileus flesh, were found to show a bilateral lamellar trama orientation (i.e. the subhymenium gelatinized but the mediostratum not so, again appearing stratose, or in Redheads terminology, bilateral; see discussion by Redhead et al., 2000).
Pileus color. European authors had described and illustrated occasional collections of "F. velutipes" with pale, sometimes white, or even variegated pileus colors. Specimens were rarely saved, but Bas (1983) proposed F. velutipes var. lactea to accommodate such forms. As our study progressed, however, it was seen that other white- or ivory-capped species existed, so while pileus color was noted for all collections, it could not be used as a species-specific segregating character.
Basidiospore size and shape. Again, Bas (1983) attempted to use spore dimensions to delineate a taxon, F. velutipes forma longispora. It was already known that F. ononidis produced long spores, but its stature and habit were quite different from those of F. velutipes. Nonetheless, spore length (especially) and shape could be used as an informative character if combined with others.
Stipe surface hyphae. In all species, the stipe is covered with a fine velvety plush, from which the common name "fuzzy foot" or "velvet shank" has come. These hyphal tips, however, are always similar to pileicystidia and somewhat boring across the genus. The character, therefore, is relatively uninformative.
Biogeographical range. It had always been assumed that the sole North Temperate representative of the genus was F. velutipes, which was widely distributed. As the study developed, however, other species with parapatric (i.e. overlapping but not congruent) geographic ranges were identified, and in the end, distribution in the Northern Hemisphere could not be used to segregate taxa. Other taxa appear to be quite limited geographically: F. mexicana to high-altitude mountain tops of central Mexico; F. stratosa to New Zealand; F. similis to Tierra del Fuego; F. callistosporioides to northern Argentina; F. ferrugineolutea to subSaharan Africa; F. mediterranea to coastal Italy; F. velutipes var. lupinicola to coastal northern California.
An interesting biogeographic puzzle was uncovered when all examined specimens from the United Kingdom proved to be F. velutipes, but almost all from Sweden were F. populicola, and those from Karsten's herbarium (Helsinki, specimens from Finland) were F. elastica. Why these relatively close regions should each be dominated by a single species, and why these species should differ over this area remain unknown. It is probable that seasonality might have sorted amongst species, with F. velutipes fruiting in winter, but F. elastica and F. populicola found somewhat earlier in the year. Moreover, with F. ononidis now known from near St. Petersburg, Russia, it would not be surprising to find that species in nearby Scandinavia.
THE TAXA
Combining the informative characters outlined above, a key to the taxa of the genus can be offered.
A. Large, prominently projecting pileicystidia entirely absent (surface always viscid in wet weather); basidiospores 11.5-15 X 7-9.8 μm; habitat arenicolous, radicating, associated with Ammophila litoralis; Italy F. mediterranea
A. Pileicystidia present, at least on pileus disc, and well-differentiated from other pileipellis elements; pleurocystidia usually present, but either rare or apparently lacking in some species or specimens;
stipe dry, usually velutinous to pruinose . .B
B. Pileus tissues strongly stratified, the upper 2/3 a thick, hyaline, geltinized layer (an overly developed hypodermium) bearing a much thinner (one tenth or less of the gelatinized layers thickness) superficial layer of clavate to fusiform cells; trama not gelatinized, exceedingly densely interwoven (lacking inter-hyphal spaces), well-delimited from the upper layer by an even more compact layer; lamellar margin sterile or nearly sterile from masses of cheilocystidia .. . C
B. Pileus with thick non-gelatinized context, bearing on it a thinner (approximately one fifth or one tenth or less of overall thickness) gelatinized layer composed of hypodermium, epicutis, and/or supracutis; lamellar margin fertile with varying densities of scattered cheilocystidia D
C. Pileipellis elements comprising clavate cells (<50%) with interspersed fusoid-ventricose pileicystidia (>50%) which tend to topple in microscopic mounts; hypodermium twice as thick as pileus context, hence pileus macroscopically
translucent-striate; basidiospores 6.3-6.5 X 3.8-4.3 μm, obovoid,
lacrymiform to broadly ellipsoid; Nothofagus in New Zealand . F. stratosa
C. Pileipellis elements virtually all ventricose pileicystidia with very few clavate cells, hence the layer a dense turf of erect cystidia; hypodermium approximately equal in thickness to trama; pileus opaque, castaneous; basidiospores 5.5-6 X 3-3.5 μm; on branches of Drimys winteri; Tierra del Fuego . F. similis
D. Pileipellis a turf or palisade forming an hymeniiform to subhymeniiform layer (a general impression; details may include a few irregular cells, including some filamentous hyphal tips), with or without apical mucronate or digitate short outgrowths . E
D. Pileipellis an ixotrichodermium of erect, filamentous, narrowly fusoid or very irregular "thorny" elements .. I
E. Hymeniiform pileipellis well-developed, composed of clavate and sphaeropedunculate cells and pileicystidia; lamellar trama either gelatinized and bilateral or hymenium lacking pleurocystidia F
E. Hymeniiform to subhymeniiform pileipellis composed of clavate cells intermixed with a few to many fusiform, strangulate-fusiform or forked ascending hyphal tips and pileicystidia; lamellar trama neither gelatinized nor bilateral . H
F. Pleurocystidia apparently absent (at least not conspicuous); lamellar trama not bilateral; a reddish brown central African species with dark yellow lamellae; not covered further .. ..F. ferrugineolutea
F. Pleurocystidia conspicuously projecting; lamellar trama bilateral, gelatinized; South and Central America G
G. Basidiospores 6.8-8 X 3.5-4.2 m m, thin-walled; stipe covering an intermittent hymeniiform layer of islands of such tissue (clavate vesciculose cells and ventricose-lageniform cystidia); lamellar trama interwoven; subhymenium not gelatinized; on Alnus jorullensis; Argentina . F. callistosporioides
G. Basidiospores 8-9 X 4.5-6.3 m m, with slightly thickened, yellowish, refractive walls; stipe covering very sparse, consisting of small groups of primarily ventricose to lageniform cystidia, not appearing as an intermittent hymeniiform layer; lamellar trama bilateral; subhymenium gelatinized, on Senecio; central Mexico . F. mexicana
H. Basidiospore mean dimensions 6-8.7 X 4-4.8 m m (Q = 1.51-1.64; Qm = 1.61), more or less ovoid, ellipsoid to amygdaliform; often associated with Populus, less so with Betula, Alnus, Pinus, Picea; Scandinavia, western North America .. F. populicola
H. Basidiospores 9.2-10.3 X 3.9-4.5 m m (Q = 2.05-2.58; Qm = 2.3), more or less elongate-ovoid, ellipsoid to amygdaliform; usually associated with Salix, less often with Populus tricocarpa, P. tremuloides, Betula, Alnus; west coast of North America, northeastern Asia, rare in eastern Europe . F. rossica
I. Lamellar trama distinctly bilateral; subhymenium gelatinized; basidiospores 6.3-8.9 X 3.2-4.1 m m (Q = 2); eastern slopes of Sierra Madre Oriental, Mexico .. F. "bilateralis"
I. Lamellar trama not bilateral, and lacking gelatinized tissue ...J
J. Basidiospores 7.2-14.8 X 3.7-6.48 m m broad (ave. = 4.52-5.4 m m broad), stout, often elongate; often on or in habitats dominated by
woody legumes in non-forested areas . .. K
J. Basidiospores narrower than 4.5 m m . M
K. Distribution Europe .. L
K. Distribution along west coast of North America; associated with Lupinus arboreus; pileus dull yellowish with cinnamon disc, up to 65 mm diam; basidiomata radicating, often in sand or caespitose on woody branches .. F. velutipes var. lupincola
L. Associated with Ononis on calcareous soil in poor grassland; pileus typically small, 5-35(-45) mm broad, honey yellow to orange brown outward, with dark orange disc; basidiomata solitary, rooting; Germany, Italy, western Russia .. F. ononidis
L. Associated with Cytisus, Ulex or other legumes; pileus not dark brown, generally paler; . variants of F. velutipes
M. Basidiospores 6-8 X 4-4.5(-5) m m, ovoid to broadly ellipsoid (Q = <2, usually 1.5-1.7); pileipellis, especially over disc, crowded with erect pileicystidia and some fusiform to narrowly clavate hyphal tips; pileus pallid, white, ivory to yellowish buff, with darker disc; usually radicate from buried wood; September;
The Netherlands, Switzerland F. fennae
M. Basidiospores usually longer, elongate-ellipsoid to cylindrical (Q = 2-3); pileipellis an ixotrichodermium of filamentous, often branched hyphal tips and pileicystidia .. N
N. Basidiospores 8-11.5 X 3-4 m m (Qm = 2.5-3);
often associated with Salix; Europe ..F. elastica
N. Basidiospores 6-9.5 X 3-4 m m (Qm = 2-2.3) O
O. Pileus yellowish to reddish brown; stipe yellowish to nearly black; cosmopolitan . . F. velutipes var. velutipes
O. Pileus and stipe ivory to nearly white and not darkening F. velutipes var. lactea
F. CALLISTOSPORIOIDES(Singer) Singer. 1964. Darwiniana 13: 183.
Ί Pseudohiatula callistosporioides Singer. 1961. Sydowia 15: 59.
Ί Cyptotrama callistosporioides Singer. 1962. Persoonia 2: 412. [nom. inval.]
Morphology: This species is represented by only a few specimens. Basidiomata are small, with pileus usually less than 2.5 cm diameter. Collected as it was in northwestern Argentina, it may remain relatively poorly understood for some time in the future. Small, wood-inhabiting mushrooms with russet to ferruginous-tan pilei are so common that the species may be easily overlooked or misidentified.
Flammulina callistosporioides finds similarities with two groups within the genus. One group, including F. stratosa and F. similis, exhibit significant gelatinized layers as part of the pileus trama. Although much thinner gelatinized layers can be found in the pileus trama of other species, in this complex it may consume up to 50% of the trama. We conjecture, therefore, that the pilei of fresh specimens must have a somewhat translucent appearance. In another taxonomic direction, lamellae of F. callistosporioides show a gelatinized subhymenium but densely interwoven, non-gelatinized mediostratum (a condition called "bilateral" by Redhead) in longitudinal sections. This arrangement is also found in F. mexicana, known only from the volcano belt of Mexico. If these two characters (i.e. significantly gelatinized pileus trama; bilateral lamellar trama) can be linked, then perhaps a continuum can be extrapolated to include F. callistosporioides, F. smilis, F. stratosa and F. mexicana. An additional taxon yet unnamed (known from a single herbarium specimen from the Sierra Madrι Oriental of Mexico) also shows bilateral lamellar trama. Together, these taxa constitute almost half of the genus, and could be separated at subgeneric rank. For a more detailed discussion and comparison, see Redhead et al. (2000).
Basidiospores (6.4-8.6 X 3.7-5 m m; Q = 1.74) are shorter than typical for the genus but not narrower, producing a Q value close to typical.
Habit: The type collection was on dead Alnus jorullensis. Whether other hosts can be exploited is unknown.
Geographic range: The species is known only from northwestern Argentina. The actual extent of range is unknown.
Mating studies: No cultures available; mating system and interspecific compatibility unknown.
Molecular biology: DNA extraction was not attempted, so placement on a phylogenetic reconstruction and RFLP signature remain unknown.
F. ELASTICA (Lasch) Redhead & Petersen. 1999. Mycotaxon 71: 293.
Ί Agaricus elastica Lasch. 1857. Bot. Zeit. 15: 171.
Morphology: In every previous case of morphological study, the best-understood morphotype was F. velutipes, and our studys bias was to separate other taxa from F. velutipes. For some collections (i.e. those of F. fennae or F. rossica) this contrast was clear, and the segregate taxa distinct. In other instances, such micromorphological distinction was subtle (i.e. F. elastica from F. velutipes). In such cases, either mating studies or molecular analyses supported or forced such distinctions, and a renewed search was made for characters which separated such morphologically similar organisms.
Lasch first recognized the organism (as Agaricus elasticus) as separate from Agaricus at that time in Tribe Collybia velutipes. In many respects, basidiomata are velutipes mimics (Fig. X), with longer spores in some collections. Pileipellis structure is similar to that found in F. velutipes (i.e. a turf of filamentous, more or less erect, often branched hyphae with occasional tapering pileicystidia) but rather different from that of F. fennae and other Flammulina species (i.e. a hymenial layer of sphaeropedunculate hyphal tips with scattered pileicystidia). Aside from sexual incompatibility, the leading characteristic of F. elastica is its longer spores [(7.5-)8-11.5(-12) X (2.3-)3-4(4.7) m m] than typical velutipes [6-9.5 X (2.5-)3-4(-5) m m]. Bas (1983) also knew of this variant, which he recognized as F. velutipes f. longispora (Bas, 1983).
Dikaryon cultures were fruited (Fig. X) on compacted sawdust, and those basidiomata yielded viable basidiospores. Fruited basidiomata, however, bore little resemblance to the dried basidiomata (and color diapositives) contributed by colleagues in Vienna (Austria) and Leiden (Netherlands). This, of course, was puzzling (as it was for other taxa as well), but extra-morphological data suggesting separation were compelling.
A long-spored form has been recognized as:
Flammulina elastica f. longispora Redhead and Petersen. 1999. Mycotaxon 71: 293.
Ί Flammulina velutipes f. longispora Bas. 1983. Persoonia 12: 62.
Habit: Basidiomata are found on wood (not apparently on soil as F. fennae and F. populicola) and seem particular fond of willow (Salix spp), although other hosts are exploited. Basidiomata may be found in autumn, but with only a paucity of material, an accurate estimation of seasonality is difficult.
Geographic range: F. elastica seems limited to western Europe, but if morphological analysis can be depended on, it would appear to be uncommon at best.
Mating studies: When SBIs of collection 9004 (Vienna, Austria) were submitted to a self-cross, a superficially bipolar mating system was revealed. As explained by Petersen et al. (1999), however, this result should not be taken as the last word, for unbalanced tetrapolarity seems to be common in the family Xerulaceae. A self-cross of Flammulina rossica produced the same result, although F. rossica and F. elastica do not share an immediate common ancestor on the phylogenetic reconstruction furnished by Hughes et al. (1999).
Petersen et al. (1999) were able to distinguish rather discrete mating groups within Flammulina which normally conformed to Redheads morphological judgements of basidiomatal identification. In the case of F. elastica, sexual incompatibility with F. velutipes was clearcut, indicating separation of the two species even though morphological differences were subtle. As noted above, separation of all other species from F. velutipes was important as the study developed. But as pairing experiments continued, and F. rossica was identified as another segregate species based on morphology, it was found that F. elastica and F. rossica were partially compatible: the two were at least related enough to recognize each other and form dikaryon mycelium which subsequently could be fruited into fertile mushrooms. Molecular data, however, did not support a close relationship: the two species were found on two distinct clades in the phylogenetic reconstruction (Hughes et al., 1999; Fig. X).
Molecular biology: ITS sequences and phylogenetic reconstruction: Hughes et al. (1999), with only three culture isolates of F. elastica in hand, and (at that time) one of F. ononidis, found that these two taxa had a common immediate ancestor, and in turn, were relatively distant from the rest of the genus (Fig. X). While they were not the most recently evolved, they were also not the most primative. The sister clade, however, contained F. populicola, F. fennae, and F. mexicana, a rather mixed bag.
RFLP signature: Using Hae III and Bst F51, F. rossica collections showed a 3-1 pattern, specific to this species. This result was more definitive than mating studies which could not reliably separate F. elastica from F. rossica.
Discussion: Flammulina elastica was first recognized from material donated from Austria, where it fruited on Salix. Subsequent material was received from The Netherlands, also on Salix. Although only a few collections have been seen, they have been grouped together by morphology and phylogenetic reconstruction. Mating studies, however, could not reliably separate F. elastica from F. rossica. It may be, of course, that these taxa can dikaryotize in vitro, but cannot do so in vivo, or that the resultant hybrids are not fit and do not survive, thus retaining the morphological uniqueness of the parents. Without more material (i.e. to identify geographic ranges more clearly) and further genetic studies, such extrapolations remains conjectural.
F. FENNAE Bas. 1983. Persoonia 12: 52-53. Figs. 1-6.
Morphology: Bas took pains, understandably, to separate F. fennae from the only two species-rank taxa known from Europe at that time, F. velutipes and F. ononidis. From F. velutipes separation was easy: 1) pileipellis structure of F. velutipes a turf of filamentous, often branched hyphal tips with scattered tapering pileicystidia, of F. fennae a hymenial layer of almost unbranched, erect hyphal tips often resembling elongate bowling pins with scattered thick-walled pileicystidia; 2) spores of F. fennae significantly smaller [(5.5-)6-7.5(-8) X (3.5-)4-4.5(-5) m m] than those of F. velutipes [(6-9.5 X (2.5-)3-4(-5) m m]; 3) F. velutipes found on wood substrates, basidiomata of F. fennae rooting perhaps to buried wood, but appearing as though from soil; 4) early autumn fruiting (September) for F. fennae, late fall, winter, early spring for F. velutipes; and 5) pileus ivory color to whitish in F. fennae, usually russet to ferrugineous brown (at least in center) in F. velutipes (although occasional pale forms may be found). Thus, careful collecting should reveal pseudorhizae.
Habit: Rooting perhaps to buried wood or woody debris, but appearing to arise from soil or humus (resembling in this character F. populicola).
Geographic range: We have seen material from The Netherlands and from Switzerland. The geographic range may be larger, but the species may be misidentified or overlooked, or it may simply fruit rarely.
Mating experiments: As explained by Petersen et al. (1999) the only available cultures identified as F. fennae came from CBS. When these cultures proved NOT to belong to this species (or even this genus), single-basidiospore isolates from a collection from Switzerland (our no. 8252; courtesy Dr. Egon Horak) were substituted. A dikaryon culture of this collection was fruited and produced additional SBIs. They were the only SBIs of the species, however. When submitted to a self-cross experiment, a tetrapolar mating system was revealed.
When "unidentified" SBIs were paired with those of the Swiss collections, all tests were incompatible. To our knowledge, F. fennae is a sexually isolated taxon.
Molecular biology: ITS sequences and phylogenetic reconstruction: In the case of F. fennae, two sources of reliable DNA were available: 1) an isotype specimen deposited at DAOM (courtesy Dr. Scott Redhead); and 2) cultures from the Swiss collection (our no. 8252). Sequences from these sources matched virtually completely.
No immediate extant ancestor for this species could be identified. Instead, the nearest relatives (3-4 ancestors distant) were F. populicola and F. mexicana.
RFLP signature: Swiss cultures of F. fennae exhibited Hae III-Bst F51 pattern 1-1. This pattern was unique to the species.
Discussion: Great confusion arose regarding cultures of F. fennae supplied by CBS (see discussion by Petersen et al., 1999). Cultures were not morphologically typical of the genus (i.e. they did not produce conidia but did produce a yellow turf of erect hyphal tips) and DNA sequences were unalignable with anything else from Flammulina. Later, cultures were established from a collection donated by Dr. Egon Horak from Switzerland. These cultures were morphologically typical of the genus, and DNA sequences were virtually identical to that derived from an isotype specimen of F. fennae at DAOM. The identity of the CBS cultures remains unknown, but several previous studies on physiology (i.e. growth rates, culture morphologies, etc.) of Flammulina species must be partially disregarded, for they used the original CBS cultures.
F. FERRUGINEOLUTEA (Beeli) Singer 1969. Beih. Nova Hedwigia 29: 93.
Morphology: Known to us only from literature [i.e. pileus reddish brown; lamellae dark yellow], the species was poorly understood by Singer when he transferred it. Illustration: Bull. Soc. Roy. Bot. Belg. 60: 155. 1928.
Habit: unknown.
Geographic range: Known only from Congo (formerly Belgian Congo).
Mating experiments: Cultures not available: mating system and interspecific compatibility unknown.
Molecular biology: DNA unavailable: placement on phylogenetic reconstruction and RFLP signature unknown.
F. MEDITERRANEA (Pacioni & Lalli) Bas & Robish. 1988. Persoonia
Morphology: Known to us only from literature, but accurately redescribed by Bas and Robich (1988). Diagnostic characters: 1) consistently viscid pileus; 2) viscid upper stipe; 3) pileicystidia absent; 4) basidiomata rooting; and 5) long spores (11.5-15 X 7-9.8 m m). These characters could suggest Xerula, Marasmius, or other genera. Illustration: Rev. Mic. Ital. 14:5. 1985.
This enigmatic species has born its share of critical examination. The species epithet was originally proposed in Hydropus (Pacioni & Lalli, 1985), presumably based on glandular caulocystidia (Robich, 1986). Two subsequent publications treated the species. Horak (1988) transferred the name to Oudemansiella, chiefly because the pileipellis was embedded in a gelatinous or viscid deposit, and the stipe was radicating. Bas and Robich (1988) treated the species as a Flammulina, citing the ixohymeniiderm arrangement of the pileipellis, paucity of caulocystidia, and apparent association with Ammophila, a grass genus in sand. They compared the radicating stipe and fruiting from buried plant debris with the same characters of F. fennae, although now other species of Flammulina are known to share this habit (Redhead and Petersen, 1999). According to Bas and Robich (1988), stipe tissue was monomitic, and explicitly not sarcodimitic, this supporting placement in Flammulina rather than in Xerula (Oudemansiella of Horak). Petersen (2000) has suggested that the species might be more comfortable in Rhizomarasmius, although this genus does not form the same type of pileipellis. Although Bas and Robich (1988) illustrated pileipellis elements and cystidial types as without clamp connections, their description cited clamps on basidia and stipe tramal hyphae.
Habit: Associated with Ammonophila litoralis in dune ecosystem.
Geographic range: Known only from the east and west coasts of Italy.
Mating experiments: No cultures available: mating system and interspecific compatibility unknown.
Molecular biology: DNA not available: placement in phylogenetic reconstruction and RFLP signature unknown.
F. MEXICANA Redhead et al. 2000. Mycologia (in press).
Morphology: Basidiomata in the field superficially closely resemble those of F. velutipes except that basidiomata appear to arise from humus at the immediate base of the shrubby dead stumps of Senecio but in reality are attached to the woody substrate just below ground level, while basidiomata of F. velutipes consistently are found directly fruiting on wood, often some distance from the ground. Microscopically, the two species are quite different. The pileipellis of F. mexicana comprises a hymeniiform layer of broadly clavate or sphaeropedunculate cells with scattered elongate, thin-walled pileicystidia. In this regard the species is similar to F. fennae and F. elastica. The gelatinous hypodermium is very thin or almost absent, but the lamellar trama is composed of a parallel mediostratum and more or less gelatinized subhymenium (called "bilateral" by Redhead). Basidiospores are more or less typical of the genus (7.1-11.7 X 4.5-6.9 m m; Q = 1.46-1.58) or slightly broader.
Habit: On dead stumps of Senecio cineraroides, a woody composite.
Geographic distribution: Known only from high altitude on slopes of Vulcan Melintzin in Estada Tlaxcala, Mexico.
Mating experiments: A self-cross of collection 6237 (type collection) revealed a tetrapolar mating system. Pairings with all other taxa of the genus were incompatible. The species appears genetically isolated.
Molecular biology: ITS sequences and phylogenetic reconstruction: Only one collection of F. mexicana yielded DNA for sequencing. That sequence was included in a clade comprising F. fennae and F. populicola. In fact, F. mexicana was nested among three collections of F. populicola. One of those F. populicola collections was from California, and this led to the conjecture that the western North American distribution of F. populicola, which extends from the Rocky Mountains to the Pacific Coast, and from Alberta to New Mexico, may have extended farther south at one time. As weather and other geophysical conditions changed, F. populicola retreated to more northern stations, but high-altitude Mexican populations became isolated. Subsequently the Mexican individual populations diverged morphologically and sexually (incompatible with all other taxa, including F. populicola) but not so much in the particular length of DNA examined by our sequencing procedure.
RFLP signature: The sole collection of F. mexicana from which DNA was extracted showed a restriction pattern (using Hae III and Bst F51) of 1-0. This pattern was shared by F. velutipes, but phylogenetic reconstruction and mating studies show that the two species are only distantly related..
Discussion: Basidiomata in the field are handsome but the species seems very limited in geographic range, as is its host, Senecio cinerarioides. That pileipellis structure resembles that found in F. populicola would not seem strange, for the sole collection of F. mexicana nests within examples of F. populicola on the phylogenetic reconstruction.
F. ONONIDIS Arnolds. 1977. Westfδl. Pilzbr. 11: 33.
= Collybia velutipes var. pratensis Schieferdecker. 1949. Zeitschr. Pilzk. 21: 21 (nom. inval).
Ί Collybia velutipes subsp. pratensis Schieferd. apud Schieferdecker & Mόller. 1963. Zeitschr. Pilzk. 29: 109. (nom. inval.)
Morphology: Basidiospores broad [7.2-13.7 X 3.7-4.49(-5.4) m m]; basidiomata usually small [pileus usually small: 5-35(-45) mm diameter], seemingly arising from soil (not explicitly on wood); stipe rooting; lamellae subdistant to distant; pileus trama not stratified; pileipellis structure of ramifying, hardly inflated hyphal tips and prominent pileicystidia (quite similar to that of F. velutipes).
Habit: Apparently associated with roots of Ononis, a genus of legumes, usually in calcareous soil.
Geographic range: Originally described from Germany and reported from Czechoslovakia, we have seen additional specimens from western Russia (north of St. Petersburg) and the Apennine Mountains of Italy (courtesy of Dr. Meinhard Moser).
Mating experiments: Until recently (Petersen et al., 1999) the only cultures of F. ononidis were deposited at CBS (whence we received their strains). Although these cultures included SBIs, a reconstituted dikaryon was fruited on sawdust and new SBIs were established from the spores of those basidiomata. A self-cross revealed a tetrapolar mating system.
When F1 SBIs from the CBS cultures were subjected to a large recognition grid, the following results showed ambiguity in sexual recognition mechanisms of F. ononidis: 1) "unidentified" strains shown morphologically to be F. velutipes (including pileus color and pileipellis construction variants) were about 50% compatible with SBIs of F. ononidis; 2) pairing of SBIs from exemplar strains (pairings = 24) also prooved more than 50% compatible with SBIs of F. velutipes; 3) in inter-exemplar pairings (n = 24) with F. populicola, F. rossica, F. elastica, and F. mexicana, complete intersterility was observed (Petersen et al., 1999).
Molecular biology: ITS sequences and phylogenetic reconstruction: At the time that DNA sequences were derived from F. ononidis, only the CBS cultures were available. This single source formed a clade with F. elastica, although relatively weakly related to the latter. This result was enigmatic considering that sexually, F. ononidis appears partially sexually compatible with only variants of F. velutipes. Nonetheless, using molecular data, F. velutipes seems only distantly related to F. ononidis. With the addition of three more collections (and cultures from them) we hope to better delineate the position of F. ononidis on the phylogenetic reconstruction (Hughes et al., 1999)..
RFLP signature: Again, only limited material was available for RFLP analysis (cultures at CBS). Hae III and Bst F51 gave a pattern of 3-0, which was unique within the genus.
Discussion: Together with F. velutipes var. lupinicola, F. ononidis seems to be associated with leguminous plants. A report from North Africa indicates that another species perhaps assignable to Flammulina occurs there also with legumes.
Morphologically, Redhead annotated many herbarium specimens as exhibiting pileipellis construction intermediate between such structures in "morphological exemplar" collections. In those cases, experiential judgement was necessary, supplemented with spore dimensions and notes on habit. Independently, pairing experiments indicated the F. ononidis was partially compatible with F. velutipes, with which it shared similar pileipellis structure but not habit.
One conjecture is that F. ononidis is a hybrid between two parental taxa of Flammulina. We have been able to pair SBIs of more recent collections of morphological F. ononidis with each other and with the original CBS SBIs, but compatibility has been consistently low; from 0-25%. This result may indicate hybrid intersterility or at least reduced compatibility, not uncommon among interspecific hybrids in green plants. Considering such partial sexual intercompatibility, the very weak relationship between F. ononidis and F. velutipes on the ITS phylogenetic reconstruction is interesting.
F. POPULICOLA Redhead & Petersen. 1999. Mycotaxon 71: 288-289.
 |
 |
 |
Morphology: As usual, as long as all segregate species had to be separated from F. velutipes, recognition of F. populicola was relatively easy. While the pileipellis of F. velutipes was dominated by slender, filamentous, often moniliform hyphal tips (plus pileicystidia), that of F. populicola was composed of a hymeniiform layer of clavate to subspheropedunculate hyphal tips (plus pileicystidia). This was combined with unusually short basidiospores (6.3-7.4 X 3.7-4.5 m m; Q = 1.62). Finally, basidiomata appeared as though they were arising from soil, but in fact were attached to buried wood, usually of poplar or aspen (Populus spp.).
Habit: On buried wood of Populus tremula in northern Europe and P. tremuloides in western North America.
Geographic range: Northern Europe, Scandinavia, Rocky Mountains and farther west in North America, as far south as Colorado and New Mexico. Recently (Gulden, 1992), F. velutipes was listed as the only Flammulina species in Sweden, but now F. populicola can be added. In fact, of the several collections we have examined from Sweden, only one was F. velutipes. Moreover, with F. elastica so common in nearby Finland, it would be amazing if that species were absent from the Swedish mycota. Moreover, we have seen a single specimen of F. ononidis from the vicinity of St. Petersburg, Russia, so it would not be surprising of it were to be found in Scandinavia as well.
Mating experiments: A self-cross of collection DED 6115 (California) revealed a tetrapolar mating system. "Recognition grids" showed F. populicola to be nearly genetically isolated from all other taxa, but rare dikaryotization events occurred with SBIs of F. velutipes, indicating that the two taxa were not yet totally genetically isolated from one another. Whether these putative "hybrid" dikaryons can be fruited remains untested, but if low level interfertility can be achieved, this would provide a mechanism to introduce genes of F. populicola into the F. velutipes genome or vice versa.
Molecular biology: ITS sequences and phylogenetic reconstruction: Three collections of F. populicola sorted into two closely related clades. One clade comprised collections from California and Sweden; the other clade represented a single collection from Alaska. Interestingly, F. mexicana (q.v.) was more closely related to the California and Swedish F. populicola collections than was the Alaska collection of F. populicola. This may support the conjecture that F. mexicana was once part of the western North American distribution of F. populicola but became isolated long ago at high elevations of central Mexico (specifically, the volcanic slopes of Estada Tlaxcala) and there diverged both morphologically and sexually.
RFLP signature: F. populicola exhibited a restriction pattern with Hae III and Bst F51 of 2-0. In previous studies using Hae II and Bgl I, restriction patterns could be used to roughly identify geographic populations, one pattern (2-2) extended from the northern Rocky Mountains into Europe, while the other pattern (2-1) was found in material from Alaska and Arizona.
Discussion: Because other workers generously donated strains of Flammulina for inclusion in our studies, it became possible to work with these strains toward their identification. Specifically, two strains donated by Dr. Hans Gruen (as F. velutipes) proved to be F. populicola. Thus a score of papers on fruiting physiology in F. velutipes require re-examination. One of those strains, however, came to Gruen from Aschen-Εberg (from Sweden), so her pioneering physiological and developmental papers must also be reinterpreted. A further ramification: one of Gruens strains (the same one which came from Aschen-Εberg) was used by McKnight and Estabrook (1986) in physiological papers on optimum edaphic conditions for fruiting of various strains of F. "velutipes," and the same strain was employed by Yokoyama in mating experiments, putatively within F. velutipes, but now recognized as including two species. Finally, the preferred strain of F. "velutipes" sold by Stamets (Fungi Perfecti) has also been identified as F. populicola.
F. ROSSICA Redhead & Petersen. 1999. Mycotaxon 71: 290.
Morphology: The first collections of this species were gathered in far eastern Russia, where local workers identified the species as F. velutipes. Pileus color, however, was pallid pinkish off-white to ivory, and the stipe showed no signs of the usual olivaceous velvet of F. velutipes. Subsequent micromorphological examination showed that the pileipellis arrangement was quite different: a hymeniiderm palisade composed of a mixture of moniliform hyphal tips quite like those seen in F. velutipes, but also broadly clavate cells whose apices often profilerated into similar narrower hyphal tips. These cell types were interspersed with the usual elongate-fusiform to lanceolate pileicystidia. As additional collections were identified, the ratio of moniliform hyphal tips to clavate cells was found to vary widely, making identification difficult in some instances. Basidiospores did not help, for they fell well within the normal range for the genus.
Habit: Substrates have not been accurately recorded, but Salix and Populus are among them. Habit is not taxonomically informative, therefore, for F. elastica seems associated with Salix, and F. populicola seems to prefer Populus.
Geographic range: Collections have been seen from coastal British Columbia, Alaska, and far eastern Russia, thus a transBeringian distribution. Cultures from the St. Petersburg (Russia) collection proved also to be F. rossica when fruited, extending the range to western Russia. In North America, the species is joined by F. populicola and F. velutipes, and F. velutipes is also known from several areas of Asia, including Japan, although not specifically from the region along the Russian coast of the Sea of Japan.
Mating experiments: A self-cross using monokaryon isolates from the type collection revealed a superficially bipolar mating system (Petersen et al., 1999). These results are open to interpretation, however, for unbalanced tetrapolarity is found in several other species of the genus. It is possible, however, that the bipolar mating system is true, for F. elastica gave the same result, and the two species are partially sexually intercompatible.
A large-scale "recognition grid" (Petersen et al., 1999) produced the following results: 1) while F. rossica and F. elastica were partially intercompatible, the rossica-elastica genetic complex was virtually isolated from all other species of the genus.
Molecular biology: ITS sequence and phylogenetic reconstruction. Given the results of mating experiments, it might have been expected that F. rossica would form a clade with F. elastica, but this was not the case. Instead, F. rossica was found in a large clade which included F. mexicana, F. populicola, and F. fennae, although relatively distantly related to all these. In fact, F. rossica was relatively isolated (Hughes et al., 1999).
RFLP signature. Using Hae III and Bst F51 enzymes, isolates of F. rossica showed a pattern of 2-1, which was unique among all tested taxa of the genus (Methven et al., 2000).
Discussion: Flammulina rossica is an enigma. Its micromorphology seems to combine features of F. velutipes with those of F. populicola, at least in pileipellis arrangement. Its substrata seem mundane and shared with other taxa. Mating experiments link it to F. elastica through superficially bipolar mating system as well as partial intercompatibility in "recognition grids" and inter-exemplar pairings, but placement on a phylogenetic reconstruction removes these two taxa from each other. That the species is not a recurring hybrid cannot be discounted. Altogether, the species presents systematics problems which have not been solved up to this writing.
F. SIMILIS Horak. 1979. Flora Cryptog. Tierra d. Fuego 11(6): 170-172.
Morphology: This species is known only from a couple collections gathered in Tierra del Fuego. Basidiomata are small (pileus probably not more than 3 cm diam); pileus trama stratified, with upper layers gelatinized, and thin lower layer densely interwoven; basidiospores small (5.5-6 X 3-3.5 m m); pileipellis minutely velutinous, an hymeniiform layer of fusoid-ventricose hyphal tips with scattered clavate cells interspersed; lamellar trama not bilateral or gelatinized.
Habit: on wood of Drymis winteri.
Geographic range: Known only from far southern Argentina, but probably also in Chilean Tierra del Fuego.
Mating experiments: No cultures have been established: mating system and intercompatibility patterns remain unknown.
Molecular biology: No DNA extraction was attempted on the small dried basidiomata: placement in phylogenetic reconstruction and RFLP signature remain unknown.
Discussion: The strongly stratified pileus trama is shared by F. stratosa from New Zealand, immediately raising the possibility that this hyphal construction type might be Gondwanan in origin. Lamellar trama of F. similis, however, does not show gelatinized subhymenium (and therefore is not "bilateral"), and so would seem less related to F. callistosporioides, also from South America, and F. mexicana. With no information from pairing experiments or molecular biology, the species remains poorly understood.
F. STRATOSA Redhead, Petersen, & Methven. 1999. Canad. J. Bot. 76: 1589-1590.
Morphology: Basidiomata are small (pileus not over 3.5 cm diam), rooting into well-rotted wood of Nothofagus; pileus trama strongly stratified, with thick hypodermium gelatinized, and lower trama densely interwoven; pileipellis hymeniiform, composed of mostly clavate to broadly clavate hyphal tips interspersed with fusoid-ventricose, thin-walled pileicystidia; basidiospores small (6.3-6.5 X 3.8-4.3 m m).
Habit: on well-rotted Nothofagus stump.
Geographic range: At present the species is known only from the South Island of New Zealand (one collection).
Mating experiments: At present SBIs are available only from the type collection. A self-cross using that collection revealed a tetrapolar mating system.
In extended "recognition grids" against "unidentified" strains of Flammulina, F. stratosa proved incompatible with all other taxa. The species seems to be genetically isolated.
Molecular biology: ITS sequence and phylogenetic reconstruction: ITS sequence from the only extant collection of F. stratosa showed the taxon to be the least related to all other taxa of the genus.
RFLP signature: F. stratosa was not included in species-specific restriction enzyme studies.
Discussion: Although basidiomata are small and slender and thus easily overlooked, the species has proven to be quite unique and interesting. First; unlike other taxa within the genus, the stipe in F. stratosa is clearly radicating, although on well-rotted wood. Basidiomata of other species (F. populicola, F. ononidis, etc.) may arise from a short stipe base attached to buried substrate, but in all these cases, the stipe is quite stout and roots only a few centimeters. Such a rooting stipe is seen more commonly in Xerula and Rhizomarasmius.
Second; pileus tramal construction is strongly stratified, with a thick hypodermium below the pileipellis, a construction shared with F. similis from southern South America. Third; ITS sequence data, although based on a single collection, indicate that the taxon is basal to the entire genus. Thus, it might be necessary to extrapolate that other micromorphological arrangements found in the genus are derived from those shown by F. stratosa (i.e. pileipellis tissues, lamellar trama construction, etc.). Fourth; such a morphologically unique, genetically isolated organism supports previous biographical conclusions concerning the long separation of taxa in Australasia, especially New Zealand.
F. VELUTIPES (Curtis: Fr.) Singer. "1949" (1951). Lilloa 22: 307.
Ί Agaricus velutipes Curtis. 1777. Flora Londinensis fasc. 4, plate 70.
Morphology: Analysis of many fresh collections and herbarium specimens yielded a core of taxonomic characters through which to diagnose F. velutipes. Pileipellis structure was composed of an ixotrichodermium of ramifying filamentous to strangulate hyphal tips with interspersed fusoid-lanceolate pileicystidia. Basidiospores were quite variable, but in "typical" material were 6-9.5 X 3-4 m m. Pileus color ranged from almost white (in var. lactea) to the usual leather colored to russet shades; stipe texture was invariably minutely velutinous, caused by a superficial layer of ventricose-lanceolate caulocystidia. Pleuro- and cheilocystidia were prominent, but not significantly different from those of other taxa within the genus. Basidiomata were always found on wood, usually away from soil (i.e. from standing trunks or branches).
Habit: Fruiting on exposed branches and trunks of hardwood trees (often on Fagus in Europe; Acer and other genera in North America).
Geographic range: Known chiefly from the Northern Hemisphere in North America, Europe and eastern Asia. Also found in temperate South America, New Zealand and southern Australia, probably from introduction by human intervention.
Mating experiments: A number of previous reports (i.e. Lamoure, 1988) concluded that F. velutipes showed a tetrapolar mating system. We confirmed this, using the designated epitype collection (Redhead & Petersen, 1999; Petersen et al., 1999).
In extended "recognition grids" against "unidentified" strains of Flammulina, the infraspecific variants of F. velutipes all proved intercompatible, forming a more or less well-defined intersterility group. About 50% compatibility was shown between all F. velutipes variants and F. ononidis, indicating their retention of sexual recognition mechanisms.
Low levels of sexual compatibility were seen between SBIs of F. populicola and the infraspecific variants of F. velutipes. Such events were much rarer in pairings between F. velutipes and all other species.
Molecular biology: ITS sequences and phylogenetic reconstruction: F. velutipes appears to form a discrete clade on a phylogenetic reconstruction. There seems to be little distinction among the infraspecific variants, although European strains seem to be grouped together. Conversely, collections from China, Canada, and United States did not form discrete groups.
One collection from Argentina, annotated by Redhead as atypical of F. velutipes but still within the taxon, was found to contain an ITS sequence composed of subsequences from two species. One parental type was F. populicola while the other was F. velutipes. Moreover, RFLP patterns indicated that the F. populicola subsequence was typical of the western North American population of F.populicola, not the population from northern Europe. To our knowledge, this is the first report of hybrid DNA within a putatively hybrid morphological entity.
RFLP signatures: Using Hae III and Bst F51, all collections of F. velutipes gave a pattern of 1-0, shared only by F. mexicana.
By adding Bgl I and Bst UI, three patterns emerged (see Materials and Methods). Pattern AD was exhibited by European collections, but also by material from Argentina, Tasmania and New Zealand. Pattern BC characterized collections from eastern Asia (China, Korea, Japan), but was also found in collections from the west coast of North America. Pattern BD was found in North American collections, but also in material from Japan, Argentina, Chile (eastern and western slopes of the temperate Andes Mountains), and Russia. It is highly plausible that commerce and emigration from the Northern Hemisphere introduced the species to the Southern Hemisphere along with exported trees, wood products, and even wooden-hulled ships.
Discussion: Redhead & Petersen (1999) found it necessary to consider Curtiss plate as a possible lectotype of the species (and therefore of its autonymic variety), but this iconolectotype could not yield data required for a modern treatment of the taxon. Upon request, several specimens of F. velutipes were gathered by English collectors (thanks to Dr. Gill Butterfills good offices) from which an epitype could be selected, represented by basidiomata, spore print, dikaryon cultures and SBIs. Thus more complete information could be gathered.
The low levels of intercompatibility between F. velutipes and F. populicola indicated the possibility of interspecific hybridization. A few of these "hybrid" dikaryons were fruited and yielded fertile basidiomata, thus supporting that such hybrids might survive in nature. If so, then taxonomic characters would be shared, diluted, and mixed in such hybrids which could, presumably, back-cross with their parents. That this is not a common occurrence is supported by the relative rarity with which such hybrids are collected.
One idiosyncracy of RFLP data indicated that while the North American signature (pattern BD) was also seen in occasional European collections, the reverse was not true (pattern AD was not found in North American specimens). Some knowledge of transAtlantic commerce over the past 300 years reveals that wood and wood products from North America were highly valued in Europe, especially oak for ship hulls, and pine for masts. Equally, the lives of Michaux, Fraser, Lyon and other 18th and 19th century botanical explorers, were spent in introducing eastern North American plants, including tree species, to European markets. While transAtlantic commerce was also active from Europe to America, wood was only rarely part of the cargo unless already finished into furniture. Moreover, even wood for furniture in eastern North America came chiefly as mahogany, satinwood and the like from the neotropics rather than from Europe.
[VAR. VELUTIPES] F. LONGISPORA Bas. 1983. 1983. Persoonia 12: 62.
This should not be mistaken for F. elastica, which also exhibits unusually long basidiospores. The only two collections we have seen which could be included as a long-spored F. velutipes showed an unusual intercompatibility pattern. These two collections were fully compatible with the variants of F. velutipes, but also with F. rossica and with F. elastica. Whether this phenomenon indicates that the long spores of these two collections of F. velutipes were inherited from an F. elastica-like parent has not been ascertained.
VAR. LACTEA (Quιl.) Bas. 1983. Persoonia 12: 63-64.
The only cultures which were reliably of this taxon came from CBS where they were deposited by Bas. As such, cultures were fruited and SBIs derived. Pileus color of the fruited basidiomata was white to ivory. SBIs proved to be completely compatible with those of other infraspecific variants within F. velutipes, as well as partially compatible with those of F. ononidis.
ITS sequences and RFLP signatures did not differ significantly from those of other infraspecific variants within F. velutipes.
VAR. LUPINICOLA Redhead et al. 1999. Mycotaxon 71: 292.
Recognized by two traits: 1) basidiospores large (7-14.8 X 3.7-6.5 m m); and 2) habit on woody lupine (L. arboreus) on the California coast. SBIs were compatible with those of other infraspecific variants of F. velutipes, and the taxon appears to be nested among other collections of F. velutipes on a phylogenetic reconstruction. The RFLP signature did not differ from that of North American F. velutipes.
The references below are only those cited in the text. More literature can be found in the bibliographies of the papers on Flammulina from the Tennessee Lab (Hughes, Petersen, Redhead, Methven as senior authors), but even those bibliographies will furnish only a fraction of the literature on this commercially important taxonomic complex.
Aschen, K. 1952. Studies on dediploidization of the basidiomycete Collybia velutipes. Svensk Bot. Tidskr. 46: 366-392.
Aschen-Εberg, K. 1958. The production of fruit bodies in Collybia velutipes. Physiol. Plant. 11: 312-328.
Arnolds, E.J.M. 1977. Einige Pilze eines Halbtrocken-rasens bei Detmold (Westfalen). Westfδl. Pilzbr. 11: 29-39.
Bas, C. 1983. Flammulina in western Europe. Persoonia 12: 51-66.
Bas, C., and G. Robich. 1988. On a false Hydropus, Flammulina mediterranea, comb. nov. Persoonia 13: 489-494.
Corner, E.J.H. 1966. A monograph of cantharelloid fungi. Ann. Bot. Mem. 2: 255 pp.
Corner, E.J.H. 1970. Supplement to "A monograph of Clavaria and allied genera." Beih. Nova Hedwigia 33: 299 pp.
Corner, E.J.H. 1991. Trogia (Basidiomycetes). Gardens Bull., Singapore, supplem. 2: 100 pp.
Corner, E.J.H. 1996. The agaric genera Marasmius, Chaetocalathus, Crinipellis, Heimiomyces, Resupinatus, Xerula, and Xerulina in Malesia. Beih. Nova Hedwigia 111: 175 pp.
Gulden, G. 1992. Flammulina. Pp. 123-124 in Hansen, L. & H. Knudsen, Eds. Nordic Macromycetes, vol. 2.
Horak, E. 1968. Synopsis generum Agaricalium (Die Gattungstypen der Agaricales). Beitr. Kryptogamen. Schweiz 13: 1-741.
Horak, E. 1988. Notizie integrative tassonomico-sistematiche su Oudemansiella mediterranea (Pacioni & Lalli, 1985, comb. nov.) Rev. Micol., Boll. Assoc. Micol. Bresadola 31: 31-37.
[Humpert, A.J., E.L. Muench, A.J. Giachini, and M.A. Castellano. 2000. Molecular phylogenetics of Ramaria and related genera: evidence from nuclear large subunit and mitochondrial small subunit rDNA sequences. Manuscript, unpubl.]
Ingold, C.T. 1980. Mycelium, oidia and sporophore initials in Flammulina velutipes. Trans. Brit. Mycol. Soc. 75: 107-118.
Jόlich, W. 1981. Higher taxa of Basidiomycetes. Biblioth. Mycol. 85: 485 pp.
Karsten, P.A. 1891. Symbolae ad Mycologiam Fennicam. Pars XXX. Meddel. Soc. Fauna Flora Fenn. 18: 61-68.
Kemp, R.F.O. 1980. Production of oidia by dikaryons of Flammulina velutipes. Trans. Brit. Mycol. Soc. 74: 557-560.
Kibby, Geoffrey (2011) Fungal Portraits No 46: Flammulina populicola and F. elastica. Field Mycology 12 (3):39-41 doi:10.1016/j.fldmyc.2011.03.003
Klan, J., D. Baudisova, and Z. Skala. 1992. Enzyme activity of mycelial cultures of saprotrophic macromycetes (Basidiomycota). III. A taxonomic application. Ceskα Mykol. 46: 85.
Kόhner, R. 1980. Les Hymιnomycθtes agaricoides. Bull. Soc. Linn. Lyon, num. spιc. 49: 1027 pp.
Lamoure, D. 1989. Species concept in the Flammulina velutipes group. Opera Bot. 100: 163-167.
McKnight, K.B., and G.F. Estabrook. 1986. Adaptations of sporocarps of the basidiomycete Flammulina velutipes (Agaricales) to lower humidity. Bot. Gaz. (Crawfordsville) 151: 528-537.
Methven, A.S., K.W. Hughes, and R.H. Petersen. 2000. Flammulina velutipes RFLP patterns identify species and show biogeographical patterns within species. Mycologia (submitted).
Pacioni, G., and G. Lalli. 1985. Entitΰ micologiche del Parco Nazionele del Circeo: XVII Hydropus mediterraneus sp. nov. Micol. Ital. 14: 5-9.
Pacioni, G., and G. Lalli. "1984" (1986). Una nuova specie di funguo psammofilo: Hydropus mediterraneus. Atti Centro Studi Fl. Mediterz. (Borgo di Taro) 6: 209-212.
Petersen, R.H. 1968. The genus Clavulinopsis in North America. Mycol. Mem. 2: 39 pp.
Petersen, R.H. 1995. Contributions of mating studies to mushroom systematics. Canad. J. Bot. 73 (supplem.): S831-S842.
[Petersen, R.H. 2000. Rhizomarasmius, a new genus in the Xerulaceae. Manuscript, unpubl.]
Petersen, R.H., and R.E. Halling.1993. Mating behavior in the Xerulaceae: Oudemansiella. Trans. Mycol. Soc. Japan 34: 409-421.
Petersen, R.H., K.W. Hughes, S.A. Redhead, N. Psurtseva, and A.S. Methven. 1999. Mating systems in the Xerulaceae (Agaricales, Basidiomycotina): Flammulina. Mycoscience 40: 411-426.
Petersen, R.H., and A.S. Methven. 1994. Mating systems in the Xerulaceae: Xerula. Canad. J. Bot. 72: 1151-1163.
Pine, E.M., D.S. Hibbett, and M.J. Donoghue. 1999. Phylogenetic relationships of cantharelloid and clavarioid Homobasidiomycetes based on mitochondrial and nuclear rDNA sequences. Mycologia 91: 944-963.
Psurtseva, N. V., and A.Y. Mnoukhina. 1998. Morphological, physiological and enzyme variability of Flammulina P. Karst. Cultures. Mikol. Fitopat. 32: 49-54.
Redhead, S.A. 1987. The Xerulaceae (Basidiomycetes), a family with sarcodimitic tissues. Canad. J. Bot. 65: 1551-1562.
Redhead, S.A., and L.L. Norvell. 1993. Notes on Bondarzewia, Heterobasidion and Pleurogala. Mycotaxon 48: 371-380.
Redhead, S.A., R.H. Petersen, and A.S. Methven. 1999. Flammulina (Agaricales): F. stratosa, a new New Zealand species distantly related to the cultivated Enoki mushroom. Canad. J. Bot. 76: 1589-1595.
Robich, G. 1986. Alcuni Hydropus delle nostre regioni. Boll. Gr. Micol. Bresadola 29: 196-202.
Yokoyama, K. 1991. Distribution and speciation in Flammulina velutipes.. pp. 198-201 in Proc. Internat. Minisympos., Res. Center, Patho. Fungi Microbial Toxicoses, Chiba Univ.