This Article Abstract Full Text (PDF) Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Kerrigan, R. W. Articles by Parra, L. A. Search for Related Content PubMed Articles by Kerrigan, R. W. Articles by Parra, L. A. Agricola Articles by Kerrigan, R. W. Articles by Parra, L. A.

Agaricus section Xanthodermatei: a phylogenetic reconstruction with commentary on taxa

     Sylvan Research, 198 Nolte Dr., Kittanning, Pennsylvania 16201, USA

Philippe Callac
Jacques Guinberteau

     INRA, MYCSA (Mycologie et sécurité des aliments) BP 81, 33883 Villenave d’Ornon cedex, France

Michael P. Challen

     Warwick HRI, University of Warwick, Wellesbourne, Warwickshire, CV35 9EF, UK

Luis A. Parra

     Avda. Padre Claret n° 7, 5° G, 09400 Aranda de Duero, Burgos, Spain

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION DISCUSSION OF THE TAXA CONCLUSIONS LITERATURE CITED

 

Agaricus section Xanthodermatei comprises a group of species allied to A. xanthodermus and generally characterized by basidiomata having phenolic odors, transiently yellowing discolorations in some parts of the basidiome, Schaeffer’s reaction negative, and mild to substantial toxicity. The section has a global distribution, while most included species have distributions restricted to regions of single continents. Using specimens and cultures from Europe, North America, and Hawaii, we analyzed DNA sequences from the ITS1+2 region of the nuclear rDNA to identify and characterize phylogenetically distinct entities and to construct a hypothesis of relationships, both among members of the section and with representative taxa from other sections of the genus. 61 sequences from affiliated taxa, plus 20 from six (or seven) other sections of Agaricus, and one Micropsalliota sequence, were evaluated under distance, maximum parsimony and maximum likelihood methods. We recognized 21 discrete entities in Xanthodermatei, including 14 established species and 7 new ones, three of which are described elsewhere. Four species from California, New Mexico, and France deserve further study before they are described. Type studies of American taxa are particularly emphasized, and a lectotype is designated for A. californicus. Section Xanthodermatei formed a single clade in most analyses, indicating that the traditional sectional characters noted above are good unifying characters that appear to have arisen only once within Agaricus. Deep divisions within the sequence-derived structure of the section could be interpreted as subsections in Xanthodermatei; however, various considerations led us to refrain from proposing new supraspecific taxa. The nearest neighbors of section Xanthodermatei are putatively in section Duploannulati.

Key words: Mushroom poisoning, nomenclature, phylogeny, systematics, toxicology

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION DISCUSSION OF THE TAXA CONCLUSIONS LITERATURE CITED

 
Agaricus L. section Xanthodermatei was created by Singer (1948) to include a distinctive group of species sharing characters with the type species, A. xanthodermus Genev. Singer in his original diagnosis characterized this section as follows: "A sectione Arvensibus Konr. & Maubl. (1924) reactione anilinica in pileo et carne intense laeteque lutea vel aurantiaca et reactione "cruciformi" (anilino-nitrico) negativa differt. Species typica: A. xanthoderma Genevier." At present, members of the group are commonly recognized by generalized biochemical characteristics of their basidiomata: (1) ‘chemical’ odors, of phenol, ink, library paste, (collectively called ‘phenolic’ or ‘carbolic’ odors) and/or of iodine or ammonia (‘iodoform’ odors), (2) the presence of context (often in the stipe base), surfaces, and/or veils that become transiently yellow when damaged by cutting, bruising or (rarely) desiccation, and more distinctly by application of KOH, (3) Schaeffer’s reaction negative (i.e. no red or orange color appears when aniline plus nitric acid are applied), and (4), typically, the presence of one or more toxic compounds such as phenol (Gill and Strauch 1984, Wood et al 1998) and/or hydroquinone (Jovel et al 1996) that can cause gastrointestinal distress progressing, in severe cases, to violent vomiting (see also De Haro et al 1999, Singer 1961, Cleland and Harris 1948). Note that a useful substitution of glacial acetic acid for nitric acid in a simplified Schäffer’s reaction has been proposed by Frank (1988).

There are some caveats to this generalization. Both odors and intensity of yellowing vary and may be quite faint in some taxa and some specimens. Rufescence, which we define as a rapid (1–5 min) development of a reddish color upon exposure or damage to some or all parts of the basidiome, in a reaction typically catalyzed by tyrosinase, is absent from most Xanthodermatei but is present in A. pseudopratensis, A. freirei, and perhaps A. laskibarii and (occasionally or obscurely) A. hondensis and A. phaeolepidotus. However, after 30–40 min most Xanthodermatei can exhibit a vinaceous to brownish discoloration of the context and surfaces, sometimes after transient yellowing on cutting or when rubbed. Whether this has in some prior reports been recorded as rufescence is unclear. Some collections of some species, particularly A. xanthodermus, are apparently eaten by some people without ill effect; however we consider any species belonging to the section Xanthodermatei not fit for human consumption because in most cases toxic effects have been observed.

This generalized combination of distinguishing characters reliably excludes other taxa from section Xanthodermatei. Species in section Arvenses Konrad & Maubl. and some species in section Spissicaules (Heinem.) Kerrigan exhibit yellowing of some parts of the basidiome, but unlike the yellowing of the Xanthodermatei, this change is usually persistent, often increasing upon drying, and is associated with almond or anise odors, a positive Schaffer’s reaction on the pileipellis (in Arvenses) or sometimes only on the stipe base or on the rhizomorphs (in Spissicaules). A few non-Xanthodermatei taxa have been associated with gastrointestinal upsets, but these appear to be idiosyncratic reactions experienced by a select few. Two species now excluded from Xanthodermatei (based on sequence data) were recorded as having either a faint phenolic odor (A. sp. RWK 1198: Kerrigan 1982) or a distinct phenol-like or iodoform odor (A. sp. CA279: Guinberteau and Callac [unpubl]).

Most recent authors, ourselves included, have long embraced Singer’s implicit hypothesis of a monophyletic, if somewhat heterogeneous, group. We investigated the hypothesis of monophyly using analysis of nuclear rDNA ITS1+2 sequences. In this paper, we present evidence for monophyly among the taxa studied, which are primarily of European and North American provenance, while excluding two other tentatively-associated species. However, some extant lineages within Xanthodermatei appear to have diverged only a short time after the sectional lineage arose. We discuss 21 included taxa, including seven new ones, three of which are described fully in companion papers (Callac and Guinberteau 2005, Callac and Mata 2005), and we provide objective, sequence based characters that extend the circumscriptions of and aid in the critical identification of the included taxa. As in the first paper in this series (Challen et al 2003), we have attempted to provide a comprehensive, though not exhaustive, analysis of taxa in the group as it occurs in the middle northern latitudes, and invite others to add further taxa to the phylogenetic framework that we have established.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION DISCUSSION OF THE TAXA CONCLUSIONS LITERATURE CITED

 
Sampling.— – Specimens of Agaricus were obtained as opportunities occurred, beginning in 1975. Cultures were preferentially prepared from explants of pileus tissue on agar-based media. In some cases, spore prints were made, or lamellae were air-dried, and cultures were made by germinating the spores on agar-based media. In most cases, after notes were taken and/or photographs were made, dried voucher specimens were deposited in herbaria. Details of the 61 unique collections from Xanthodermatei affiliated taxa and 20 outgroup collections from other sections of Agaricus subgenus Agaricus evaluated in this study are summarized in TABLE I. The sequence of Micropsalliota sp. TL6025 (GenBank AF482835 [GenBank] ; Vellinga et al [2003]) was evaluated as outgroup in some analyses.

Phylogenetic analyses.— – The resulting ITS data sets were evaluated using three alternative tree-building methodologies: (i) Distance (Neighbor-Joining, or N-J) trees were produced from corrected ClustalW alignments within MegAlign; (ii) The maximum parsimony (MP) criterion in PAUP* (v 4.0b10; Swofford 2000) was used for the extensive evaluation of large numbers of trees (1–10 M per run) (iii) Quartet puzzling trees (n = 25 000) with maximum likelihood (ML) branch lengths were produced using TreePuzzle (v5.0; Schmidt et al 2000) and the HKY substitution model (Hasegawa et al 1985) for a unique sequences dataset (56 Agaricus sequences). MP heuristic search settings were as follows: random addition sequence, one tree held at each step during stepwise addition using the tree-bisection-reconnection algorithm, branch collapsing if maximum branch length was zero, gaps treated as ‘missing’, stepwise descent option not in effect, topological constraints not enforced, bootstrapping (n = 10 000) and maxtrees = 100. Ambiguous regions of the alignment were excluded from the MP analyses.

Nucleotide sequence and MSF accession numbers.— – Novel ITS1+2 sequences used in this study are available within the EMBL/GenBank databases under accession numbers DQ182509 [GenBank] -34 and DQ185550 [GenBank] -72. The 736-character MSF alignment file was deposited in the EMBL-Align database as accession ALIGN_000936.

Taxa, circumscription and commentary.— – The GeneDoc package (Nicholas et al 1997) was used to identify diagnostic ITS polymorphisms useful in circumscribing included taxa, with and without the 20 outgroup species. Positional data were derived from an MSF comprising 736 characters (81 Agaricus ITS1+2 sequences). Polymorphic characters, with flanking sequences, unique to all representatives of an individual taxon are indicated in bold uppercase type. Polymorphisms that differentiate taxa only within section Xanthodermatei are shown in normal uppercase type. Characteristic deletions are shown with an integer representing their length, i.e. (-n-). Heteromorphic characters are also shown in parentheses; these putatively allelic polymorphisms segregate (or are expected to) among relevant isolates. It should be understood that the ‘uniqueness’ of the apomorphic characters useful in taxon identification is established within the context of currently available data. As more taxa are sequenced it would not be surprising to find that a few of the included characters are plesiomorphic or homoplasic with respect to those of other taxa in the genus or section.

Nomenclature.— – The correct names (in italics and boldfaced type) have been arranged in alphabetical order. For each name a full and direct reference to its author, place of publication with page, plate or name heading number (when no page number is printed) and date are given. Author abbreviations have been taken from Kirk & Ansell (1992), publication abbreviations from Lawrence et al (=B-P-H; 1968) and Stafleu & Cowan (=TL-2; 1976–1985) and herbarium acronyms from Holmgren et al (1990). For those works which are not compiled and abbreviated in B-P-H and TL-2 the entire title of work is usually used.

If necessary, after the full reference the following data, in square brackets, have been provided:

1. In quotation marks, the year appearing in the printed matter when different from the year of actual publication.
   
2. In quotation marks, the original spelling of the taxon name when different from the correct name.
   

Under the correct name, synonymous and other names have been arranged as follows:

1. Homotypic synonymous (‘’ symbol) in chronological order.
   
2. Heterotypic synonymous (‘=’ symbol) in chronological order.
   
3. Doubtful synonymous (‘?’ symbol) in chronological order.
   
4. Misapplied and invalid names (‘–’ symbol) in chronological order.
   

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION DISCUSSION OF THE TAXA CONCLUSIONS LITERATURE CITED

 
Phylogeny reconstruction.— – Distance method.. Two Neighbor-Joining trees were produced with or without the inclusion of a sequence from a Micropsalliota species. The alignment comprising 81 Agaricus sequences, including all Xanthodermatei sequences plus 20 additional sequences representing six or seven other sectional-level groups, produced a tree in which Xanthodermatei was monophyletic (FIG. 1). Within this section, two sub-groups with relatively long branches arise from the most basal nodes of the section. The first, relatively most basal lineage includes the species A. hondensis, A. freirei and A. phaeolepidotus. These are erect, sylvan species with weak or spatially restricted yellowing reactions and some tendencies toward rufescence. The second sub-group comprises A. pseudopratensis, a short-statured species with a weak odor and with weak yellow staining that may be obscured by a prompt rufescent reaction. The two (or possibly three) lineages formed by these four species arise virtually at the sectional boundary and arguably represent subsections; however their morphological and biochemical features are somewhat variable and as they branch near the sectional origin we have refrained from proposing new ranked taxa for them. The remainder of the section comprises a grade-structured sub-group with increasingly short branch lengths at increasingly distal nodes, implying that relatively more recent speciation events are associated with greater diversification among extant species. This third lineage comprises the more ‘classic’ Xanthodermatei species, generally with more pronounced odors and yellowing reactions, and an absence of rufescence (but see A. laskibarii). With respect to inter-sectional relationships in this tree, Duploannulati was sister to Xanthodermatei, while Spissicaules was most distant.

View larger version (17K): [in this window] [in a new window]   FIG. 1. Phenogram of distance relationships (Neighbor-Joining algorithm) among 81 Agaricus ITS1+2 sequences. Collection and isolate data are given in TABLE I. Arrow indicates section Xanthodermatei. Agaricus bresadolanus represents section Spissicaules.

A second distance analysis using DNADist, with only Agaricus species and with A. pattersonae RWK 1415 as outgroup, returned essentially the same topology as the Neighbor-Joining method (not shown).

Parsimony method.— – One M to 10 M trees were evaluated in each run. In one MP analysis of 10 M trees, 1974 equally most parsimonious trees were obtained. Strict consensus trees produced from all MP trees resulting from the alignment of 81 Agaricus sequences always included three primary clades: (1) Xanthodermatei, (2) Duploannulati, and (3) all other species and sections (FIGS. 2, 3). This was true whether coded gap data was included or not; when gaps were included the lengths of the most parsimonious trees increased from 558 to 632 steps, but consensus tree topology was unchanged. Within Xanthodermatei, four clades arose from the most basal nodes in the following basal > distal order: (1) A. freirei and A. hondensis, (2) A. phaeolepidotus, (3) A. pseudopratensis, and (4) all remaining species, which formed the most distal clade with a multi-level structure. Based on branch lengths from the hypothetical common ancestor, this latter clade (the ‘core’ or ‘classical’ Xanthodermatei group) appears to have undergone more extensive sequence evolution than have those taxa arising from the more basal nodes within the section. Except for this latter point, tree topology was very similar to that obtained from this dataset using a distance method.

View larger version (25K): [in this window] [in a new window]   FIG. 2. One of 1974 equally most parsimonious trees out of 10 M evaluated. Arrow indicates section Xanthodermatei. Bootstrap values obtained in a separate parsimony analysis of this data set are shown, when at least 50%, for the relevant branches.

View larger version (25K): [in this window] [in a new window]   FIG. 3. Strict (100%) consensus of 1974 equally most parsimonious trees. Branches shown are present in all trees. Arrow indicates section Xanthodermatei.

When the Micropsalliota sp. sequence TL6025 was introduced as outgroup, consensus trees, including strict consensi of most parsimonious trees, again exhibited basal polytomies. Xanthodermatei became paraphyletic with several co-equal sub-lineages, one of which was Duploannulati. This is reminiscent of the effect of sequence TL6025 on Neighbor-Joining trees.

Maximum likelihood method.— – Using TreePuzzle with 25 000 quartet-puzzling steps, on a 56-sequence Agaricus dataset excluding identical sequences, and with A. pattersonae RWK 1415 as outgroup, a tree topology generally corresponding to those from other methods was obtained (FIG. 4). Duploannulati was again proximate to Xanthodermatei, which was again divided into the three ‘subsectional-level’ clades. Support for a monophyletic Xanthodermatei, however, was only 54%, probably reflecting the close relationship of section Duploannulati as well as the rather distinctive nature of the sequences of the hondensis-freirei-phaeolepidotus lineage. Selecting other non-Xanthodermatei species as outgroups did not have a substantial effect on ML tree topology.

View larger version (24K): [in this window] [in a new window]   FIG. 4. Maximum likelihood tree (Quartet-Puzzling algorithm) of 56 unique sequences drawn from dataset of 81 ITS1+2 sequences. Arrow indicates section Xanthodermatei. Support values, when at least 50%, are shown for the relevant branches.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION DISCUSSION OF THE TAXA CONCLUSIONS LITERATURE CITED

More pertinently, the phylogenetic signal at basal nodes is degraded by the discarding of data that occurs in the bootstrap, and under this condition monophyly of Xanthodermatei is supported only about 50% of the time, or less. This ambiguous result reflects our original sense of the section based on morphological and biochemical observations: there is significant heterogeneity within a relatively unified group of species. Furthermore, ITS1+2 sequences of Xanthodermatei are not profoundly different from those of the putative sister section Duploannulati. Unlike the case of section Duploannulati (see Challen et al 2003) there are no ITS1+2 polymorphisms that uniquely define Xanthodermatei. Briefly put, support for a monophyletic Xanthodermatei based on ITS1+2 sequences is broad but not deep.

If the established concept of Xanthodermatei is employed, all species with the combination of a phenolic &/or iodoform odor, transient yellowing (sometimes induced by KOH), and (when known) toxicity—as in all species we have studied—are included. However, two (or three) basal lineages within Xanthodermatei have an extended history of evolutionary independence from the ‘classical’ species. Three erect, sylvan species with an aspect more like some members of Sanguinolenti form one distinct lineage (or, in some analyses, two adjacent clades). One of these species, A. hondensis, either agrees with or is very close to A. H. Smith’s early concept of A. sylvaticus Schaeff. (Smith 1940, 1944, 1958, 1971; since modified), a species which is routinely placed in Sanguinolenti. Agaricus hondensis has a felt-like annulus not characteristic of ‘classical’ Xanthodermatei such as A. xanthodermus. Odor and yellow reactions are mild to obscure everywhere except at the base of the stipe in this species. It also has relatively high levels of hydroquinone (Jovel et al 1996). Whether this might signal biochemical divergence between putatively subsectional lineages requires further study. Agaricus freirei, with prompt rufescence that is unusual in the section, and also with a felt-like annulus, was also initially placed in Sanguinolenti. The preponderance of evidence indicates that these three species (including A. phaeolepidotus) belong to a clade that diverged shortly after the presumed Xanthodermatei–Duploannulati split. Six shared sequence characters mark this clade.

Another quasi-basal lineage within Xanthodermatei is represented by A. pseudopratensis. This species also has a weak (in our experience) phenolic odor, and flavescence that is often obscured by a prompt, moderate rufescence that is not characteristic of most other Xanthodermatei. This lineage seems fairly distinctive, but as only one species is known to belong to it, it is not possible to generalize further.

The remaining species in our sample form a single clade in all analyses. These ‘classical’ Xanthodermatei include the type species, A. xanthodermus. Within the section, this is the most species-rich lineage in the northern temperate zone; speciation appears to be continuing among various included taxa.

Most Northern Hemisphere species in all lineages within Xanthodermatei appear to be restricted to continental (or smaller) regions (i.e. Europe vs North America) with the exception of A. xanthodermus, A. iodosmus and perhaps the more tropical species A. endoxanthus and A. rotalis, if the latter two species should be found to intergrade. This is strikingly different from section Duploannulati, in which most species have been found in both North America and Europe (Challen et al 2003).

	DISCUSSION OF THE TAXA

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION DISCUSSION OF THE TAXA CONCLUSIONS LITERATURE CITED

 
Section Xanthodermatei Singer 1948. Sydowia 2:36.

Characteristics of the section: Basidiomes at least weakly or latently flavescent, occasionally rufescent, odor weakly to strongly ‘phenolic’ or iodoform, KOH and NaOH (on surfaces and/or context) yellow, aniline yellow to orange, aniline plus nitric acid (Schäffer’s or ‘cross’ reaction) negative (not orange or red), o-tolidine blue, toxic compounds generally present, annulus present, pendant or essentially so, cheilocystidia often present, spores lacking a germ pore. ITS1+2 sequence: no unique unifying characters are present.

Included taxa.— – Agaricus californicus Peck 1895. Bull. Torrey Bot. Club 22:203.

= Agaricus subnitens Peck 1909. Bull. Torrey Bot. Club 36:335.

= Agaricus bivelatus Peck 1909. Bull. Torrey Bot. Club 36:335.

Characteristic ITS polymorphisms.. tcagcTtttat @ 145

Discussion.. Agaricus californicus is an ecologically versatile fungus, occurring frequently in lawns (where it can be mistaken for A. campestris L. : Fr., especially by persons unable to detect its mildly phenolic, ‘library-paste-like’ odor) and less frequently under live oaks, in young cypress groves, and in landscaped areas. It may occasionally colonize garden composts and can be cultivated. While common in western California, its range is not well documented. The types of A. californicus, A. subnitens, and A. bivelatus were all collected in the eastern Los Angeles basin. We (Callac and Mata 2005) have made one collection of A. californicus in southern Mexico. It has caused some cases of gastrointestinal distress (Kerrigan unpubl data). Freeman (1979) considered this taxon to be a nomen dubium, but for us it is a good, well-known species.

The type of A. californicus (NYS: McClatchie 849) is heterogeneous (the box label reads "A. J. McClatchie Jan 10 ‘95 [/] & (W. R. Dudley fragments) [/] Type from Pasadena". Evidently Peck added later material from Dudley to the type. Within the box were found approximately five fragmented basidiomata (one stipe was in a plastic sleeve labeled "annular tissue" in A. E. Freeman’s hand) plus four unlabeled paper packets containing additional fragments of basidiomata. It seems likely, though not certain, that the packets contain Dudley’s material. The loose material in the box agreed with Peck’s description and with observations on recent collections (Kerrigan 1986). Because the original construction of the four packets may not have completely restricted movement of material in and out of the packets, RWK enclosed each in a sealed plastic bag after designating them A, B, C, and D, respectively. Elements A, C, and D also agreed with the concept of A. californicus, for example spores ca. 5.1–6.4 x 4.1–5.1 µm (Peck 1895). However, element B has larger spores, ca. 7.8 x 5.4 µm on average, or "6.1–9.2 x 4.6–6.1" µm per Freeman’s annotation. Lectotypification of A. californicus is therefore in order. The stipe sleeved by Freeman and an almost entire pileus, found among the loose material, that fits perfectly to that stipe were segregated into a fifth plastic bag labeled "PACKET E" by RWK.

Lectotype (here designated): USA, California, Los Angeles County: McClatchie 849, single basidiocarp in two pieces in plastic bag labeled "PACKET E" (NYS!).

The type of A. subnitens (NYS: 5124 C. F. Baker Jan 1909) had spores somewhat smaller than reported by Peck (RWK: 5.6–5.7 x 4.5–4.6 µm, vs. "6–8 x 4–5" µm: Peck 1909). The type of A. bivelatus (NYS: 5156 C. F. Baker Jan 1909) had spores of the size reported by Peck (RWK: 5.7 x 4.4 µm; Peck (1909): "5–6 x 4–5" µm). The specimens of both taxa, accompanied by good notes and sketches, agree with A. californicus (Peck’s notes on A. bivelatus were found with the type of A. subnitens). An isotype of A. bivelatus (5156 Claremont, Calif. Jan, 1909. Coll. Baker [BPI]) also agreed with the type and with A. californicus in all features including spore size.

Agaricus endoxanthus Berk. & Broome 1871. J. Linn. Soc., Bot. 11:548.

= Psalliota endoxantha (Berk. & Broome) Petch 1917. Ann. Roy. Bot. Gard. (Peradeniya) 6:320.

Characteristic ITS polymorphisms.. tggtcGgctcc @ 559

Discussion.. Agaricus endoxanthus is a tropical species originally recorded from Sri Lanka, but also later cited from tropical countries of Africa, Asia and Central and South America, which we have several times collected on humus of a tropical greenhouse in Spain (Parra 2002). On account of its slender habit, color of the pileus cuticle, and intense yellow change of the context at stipe base this species resembles A. moelleri, which can be distinguished by its pileipellis, constituted of filamentose hyphae not or slightly narrowed at the septa and sometimes with diffuse intracellular pale brown pigment, which in A. endoxanthus is, in contrast, constituted of terminal catenulate dolii-form elements filled with dark brown vacuolar content.

Agaricus rotalis, recently described as a new species (Peterson et al 2000; see below) is closely related to A. endoxanthus. Microscopical features from Spanish material as well as from the type of A. endoxanthus (examined by LAP) fit very well with those of Agaricus rotalis (TABLE II.).

Although by using the traditional macro and microscopical characters A. rotalis and A. endoxanthus are practically indistinguishable, they are evidently separable by their ITS sequence data. The available ITS1+2 sequences of the two species differ by three characters; two additional heteromorphisms are present. There is some evidence to suggest that sequences among isolated island populations, such as in Hawaii, may diverge relatively more rapidly than in continental populations (Kerrigan 2005; Johnson and Seger 2001).

Agaricus freirei Blanco-Dios 2001. Doc. Mycol. 31(121):28.

Key synonyms.. none.

Characteristic ITS polymorphisms.. ttgaaCtatgt @ 19; gcctgCcagag @ 236; ttctaCgccta @ 302; gaaaaAttgta @ 314; cttgtCgaggt @ 551

Discussion.. Agaricus freirei was described from material collected on coastal dunes in Spain, Pontevedra, O Grove, San Vicente do Mar, Meixilloneira beach. Parra collected it in Portugal (LAPAG 189); it was also collected along the French Atlantic littoral (CA186, 187, 202, 203) generally on sandy calcareous soil either in the Pino-Quercetum ilicis habitat where it is frequent, or, less frequently, under Cupressus macrocarpa where it has been previously inventoried in seven sites but determined at this time as A. phaeolepidotus (Guinberteau et al 1998). [In the same region (southwestern France), A. phaelolepidotus is in fact a rare species that we have collected only in three sites far from the littoral.] We also report one coastal Mediterranean collection (CA296, Porquerolle Island, France). This species was originally described as belonging to the section Sanguinolenti, but its positive yellow KOH reaction and its phenolic or iodine-like odor as well as its DNA sequence place it within the section Xanthodermatei. This medium to large species (cap diameter: 6–13 cm) is morphologically very similar to A. phaeolepidotus, and, like it, has a faint phenolic odor and does not clearly exhibit the truncated conical pileus that characterizes most species in section Xanthodermatei. Agaricus freirei can be distinguished from A. phaeolepidotus by its characteristic traits: faint to absent yellowish discoloring of the flesh of the stipe quickly followed by a red vinaceous discoloring; cap color varying from mahogany brown to chocolate brown; annulus pendant, membranous, simple with a thicker double margin, the lower one being brown and breaking up to form an interrupted circle of narrow arc-shaped squamae; more robust habit with infrequent rhizomorphs; smaller and less conspicuous cheilocystidia. In other respects, A. freirei closely resembles A. hondensis, which, based on ITS1+2 sequence data, may be its closest relative. The three species form a basal clade in distance and ML analyses, otherwise a pair of sister lineages, within Xanthodermatei.

Agaricus hondensis Murrill 1912. Mycologia 4:296.

= Agaricus bivelatoides Murrill 1912. Mycologia 4:297.

= Agaricus hillii Murrill 1912. Mycologia 4:298. ["Hillii"]

= Agaricus mcmurphyi Murrill 1912. Mycologia 4:299. ["McMurphyi"]

= Agaricus glaber Zeller 1938. Mycologia 30:469. ["glabrus"]

= ?Agaricus abramsii Murrill 1912. Mycologia 4:298. ["Abramsii"]

= ??Agaricus subrufescentoides Murrill 1912. Mycologia 4:299.

–Agaricus sylvaticus Schaeff. sensu A.H. Smith (pre 1975).

Characteristic ITS polymorphisms.. tttgcTtgtca @ 233; ttgagAtcagc @ 556; gctctTagtgt @ 644

Discussion.. Agaricus hondensis, described from material collected under Sequoia in La Honda, San Mateo Co., California, USA, is a characteristic if not superabundant fungus of the Pacific North American temperate forests. It is one of relatively few Agaricus species found under redwoods; in California it is also occasionally found with Cupressus macrocarpa and C. abramsii. North of California it occurs with other conifers of the coastal region, and it may occasionally be found under hardwoods and in the Sierra Nevada mountains. It is typically a tall, robust mushroom emerging from deep needle litter. The pileus color varies from cream-color to deep vinaceous brown. In its darker color phase it resembles a sister species from Europe, A. freirei. A. hondensis can cause vomiting and gastrointestinal distress (Smith 1975, Jovel et al 1996, Kerrigan unpubl data).

Murrill (1912) collected and described six species with generally similar aspects and spore sizes. Synonymy among at least some of these species seems even more plausible given that spores of the type of A. hondensis (NY: W. A. Murrill and L. R. Abrams 1260) are somewhat larger than reported by Murrill ("5 x 2.5" µm: Murrill 1912): 5.0–5.3 x 3.7–3.9 µm, measured from lamellae by RWK. This is in better agreement with our concept of the species: 4.7–5.7 x 3.3–3.5 µm (spores from lamellae may deviate somewhat from norms). Smith (1944) and Isaacs (1963) placed three of Murrill’s 1912 species, A. bivelatoides, A. hillii, and A. mcmurphyi, and also A. glaber Zeller, in synonymy with A. hondensis. Our own (RWK) examinations of the types of these four species [all at NY] do not contradict Smith and Isaacs (the type of A. mcmurphyi is too deteriorated to provide morphological data). However the truncate-conic shape of the pileus of A. bivelatoides as described and sketched is not typical of A. hondensis, which is normally convex until plane. The pleasant taste and odor recorded for A. hillii (Murrill 1912) might be explained by the fact that the phenolic odor of A. hondensis is relatively mild and is largely restricted to tissue at the base of the stipe; also, in our experience, a number of people have difficulty detecting phenolic odors in Agaricus. Agaricus abramsii Murrill is based on a distorted specimen in poor condition (Smith 1944: spores 5–6 x 3–3.5 µm) and was not critically studied by us.

Agaricus subrufescentoides Murrill has the aspect of (brown) A. hondensis; a watercolor of the type shows no distinctive differences (or conclusive similarities). Both the type from Seattle, Washington (W. A. Murrill 591[NY]) and the other authenticated collection (W. A. Murrill 720, Tacoma, Washington [NY]) have somewhat longer spores than typical A. hondensis (RWK: 6.0–6.1 x 4.0 µm and 6.2–6.4 x 4.4–4.5 µm, respectively, or 5–6.3 x 3.3–3.8 µm for the type per Smith [1944]). The described rufescence of A. subrufescentoides might argue against synonymy with A. hondensis, in which we have observed only slow, vinaceous discoloration (and possibly a very obscure sordid ‘rufescence’). Smith (1944) regarded this species to be identical to his concept of A. sylvaticus Schaeff., but later (Smith 1975) acknowledged A. sylvaticus to be a different species and treated Pacific coast material as A. hondensis (however see also Smith et al 1979). Possibly complicating this picture is Smith’s (1940) description of A. sylvaticus from the west coast of the USA as having a ‘slightly amygdaline taste’ and ‘no readily discernible [rapid] color change to red’ in basidiomata. Brown capped, longer-spored North American members of this group deserve more study; until then their status will remain unclear.

Agaricus iodosmus Heinem. 1965. Bull. Soc. Mycol. France 81(3):399.

= Agaricus pilatianus (Bohus) Bohus 1974. Ann. Hist.-Nat. Mus. Natl. Hung. 66:78.

= Agaricus pilatianus f. magnus Bohus 1974. Ann. Hist.-Nat. Mus. Natl. Hung. 66:79.

= Agaricus pilatianus f. silvaticoides Bohus 1974. Ann. Hist.-Nat. Mus. Natl. Hung. 66:80.

= Agaricus pilatianus var. iodoformicus Hlaváek 2001. Mykol. Sborn. 78(3–4):115.

Characteristic ITS polymorphisms.. ctgttTttgaa @ 275; aaattAtaata @ 317; gtcttTagtga @ 671; taggaT-taccc @ 720

Discussion.. Agaricus iodosmus, described from material collected in a garden in Rabat, Morocco, is an uncommon Agaricus species cited from Southern Europe, principally in the Mediterranean area, from North Africa and now from New Mexico (USA). Macroscopically, this species is characterized by its robust to semi-squat habit, its strong chrome-yellow discoloration on the pileus margin, annulus and lower stipe when rubbed, and its pendant annulus recurved against the stipe, showing three edges delimitating two concave unbroken collars. Microscopically, it is characterized by its two types of cheilocystidia, one simple and claviform, the other multiseptate and very conspicuous, sometimes with refractive walls, their proportion being very variable.

Agaricus pilatianus is here considered as a synonym of A. iodosmus. According to the original descriptions these two species can be separated by their odor, size of pileus and coloration of pileus surface. Agaricus pilatianus was described with a "carbolic acid" (phenolic) odor and a pileus surface changing from white in young specimens to brown-greyish in mature specimens, and measuring 6–12 cm diam.; A. iodosmus was described with an iodoform-like odor and darker pileus surface (sooty-brown already when young) measuring 12–20 cm diam. However, these features are very variable and considered by us to be without taxonomic value in this instance, since intermediate specimens between both species concepts can be observed in nature, and found in the literature. In fact LAPAG 203 had a dark pileipellis and phenolic odor, LAPAG 238 had a dark pileipellis and iodoform-like odor and LAPAG 245 had a paler greyish pileipellis and phenolic odor.

Bohus himself (1974) admitted a great variability as far as the color and the squamosity are concerned, describing two new forms from A. pilatianus, called form sylvaticoides "much resembling A. sylvaticus Schaeff." and form magnus "pileus 13–18 cm diam at first entirely brownish grey, resembling A. iodosmus but with an other smell." On the other hand, Hlaváek (2001) recently has described an A. pilatianus var. iodoformicus, that is an A. pilatianus but with an iodoform-like odor. Furthermore, a single specimen can have a phenolic odor for some persons and an iodine-like (betadine) one for others.

Agaricus laskibarii L.A. Parra & Arrillaga 2002. Doc. Mycol. 31(124):34.

Key synonyms.. none.

Characteristic ITS polymorphisms.. tttcgTtggat @ 161; tctaa(t/C)cgtct @ 664

Discussion.. Agaricus laskibarii, described from material collected on sandy soil from coastal dunes in France, Landes de Gascogne, Seignosse, is a very rare Agaricus species that has not been found since its original description. Macroscopically, this species closely resembles Agaricus xanthodermus at first sight, because of its slender habit and annulus structure, but when bruised at the stipe base, this latter species gets a stronger yellow tinge while the context of Agaricus laskibarii becomes dirty rosaceous in 2–3 minutes after a transient very slightly yellow discoloration. The description of Agaricus laskibarii, although based on fresh basidiomata, was done one day after collecting; probably the overall character of color changes of the context are still not sufficiently known in this species to consider this a definite rufescent reaction. Microscopically, A. laskibarii differs from all other Xanthodermatei by its single, often flexuous or constricted, versiform (cylindrical, clavate, fusiform, lageniform, ampulliform or utriform often capitate or rostrate) cheilocystidia.

Agaricus menieri Bon 1981. Doc. Mycol. 11(44):28. [nom. nov. based on Psalliota ammophila Ménier]

= Psalliota ammophila Ménier 1893. Bull. Soc. Sci. Nat. Ouest France, 3(2):67.

= Agaricus ammophilus (Ménier) Sacc. 1895. Syll. Fung. 11(3):69. [nom. illeg.; art. 53.1 non Agaricus ammophilus Durieu & Lév. 1868. Expl. Sci. Algérie:14. qui est Psathyrella ammophila (Durieu & Lév.) P.D. Orton 1960. Trans. Brit. Mycol. Soc. 43(2):160.]

Characteristic ITS polymorphim.. tgggcCttcta @ 296; tgaaa(-2-)ttgta @ 313; ggccgTttgat @ 546

Discussion.. Agaricus menieri is a large and robust species typically growing in the Atlantic or Mediterranean coastal dunes (France, Spain, Portugal, Italy, Israel and, according to Wasser (2002), the British Isles) where it can be locally abundant. An exceptional collection (LAPAG 237) was made in Aranda de Duero, Spain, more than 230 km from the coast (Parra 2003). Until the cap opens, the sporocarp is often hypogeous and entirely chalk-white colored. The base of the stipe is generally claviform and its yellow discoloring by handling or cutting is strong and immediate. We have always found a strong phenolic odor without any anise component in agreement with Béguet (1967) and we believe that the odor has been initially misinterpreted as aniseed (Ménier, 1893). With its narrow, typically fleshy annulus adpressed to the stipe, Agaricus menieri can be easily distinguished from several related species (e.g. A. xanthodermus, A. moelleri, and A. placomyces) but not from A. iodosmus, which mainly differs by its browngrey cap. Agaricus menieri is also characterized by the largest spore size (7–9 x 5–5.5 µm) in section Xanthodermatei.

Agaricus moelleri Wasser 1976. Novosti Sist. Nizsh. Rast. 13:77.

= Agaricus xanthodermus var. obscuratus Maire 1910. Bull. Soc. Mycol. France 26:192.

= Psalliota meleagris Jul. Schäff. 1925. Z. Pilzk. 4(2)(Neue Folge):28.

Agaricus meleagris (Jul. Schäff.) Imbach 1946. Mitt. Naturf. Ges. Luzern 15:68 [nom. illeg. non A. meleagris Sowerby 1798, Col. Fig. Engl. Fung.:pl.171 qui est Leucoagaricus meleagris (Sowerby) Singer 1951. Lilloa 22: 422. ["1949"]]

Agaricus praeclaresquamosus Freeman 1979. Mycotaxon 8(1):90. [nom. nov. based on Psalliota meleagris Jul. Schäff. 1925]

= Psalliota meleagris var. terricolor F. H. Møller, Friesia 4(3):208. 1952 ["1951"]

–A. placomyces auct. eur. plur. non Peck 1878, New York State Mus. Ann. Rep. 29:40.

Characteristic ITS polymorphisms.. none (but see discussion of A. xanthodermus).

Discussion.. Agaricus moelleri was described from material collected in Denmark under fruit trees on bare soil. The species is widespread in Europe. Based on study of herbarium material, Freeman (1979) believed it to occur "from Maryland to Florida." We have not been able to confirm this. Kerrigan (1982, 1986) reported it (as A. praeclaresquamosus) from California, but further study has shown that this material differs morphologically and, in at least two instances (see A. sp. RWK 1937 and A. sp. ML5, discussed below) also differs at the DNA level. The European species Psalliota meleagris Jul. Schäff. (1925) was transferred to the genus Agaricus by Imbach in 1946. Shortly after, Pilát (1951) pointed out that the American Agaricus placomyces Peck could be a synonym with Agaricus meleagris (Jul. Schäff.) Imbach owing to their macro- and microscopic similarity. Later, Dennis (1960) realized that A. meleagris (Jul. Schäff.) Imbach is a later homonym of Agaricus meleagris Sowerby (1798), but he did not create a new name for it because he considered A. meleagris (Jul. Schäff.) Imbach as a synonym of A. placomyces which would be the correct name for a combined species. So, from 1960 the name A. placomyces was broadly used by European mycologists (for instance Bohus 1974, Phillips 1981, Michael et al 1983) referring to A. meleagris (Jul. Schäff.) Imbach in synonymy.

In 1984, Cappelli published his monograph on European Agaricus where he considered A. placomyces different from A. meleagris (Jul. Schäff.) Imbach adopting for this latter the name A. praeclaresquamosus A. E. Freeman, a new name for A. meleagris (Jul. Schäff.) Imbach created by Freeman (1979), who also considered this species different from A. placomyces. Cappelli included A. placomyces auct. europ. non Peck under A. praeclaresquamosus synonymies, since then being unanimously followed by Agaricus monographers (Bon 1985, Meusers 1986, Bohus 1990, Bohus 1995, Wasser 2002) who do not consider A. placomyces a synonym with A. praeclaresquamosus (= A. meleagris (Jul. Schäff.) Imbach).

Given the foregoing we believe that the epithet A. placomyces was probably considered as a synonym with Agaricus meleagris by European mycologists from Dennis (1960) to Cappelli (1984) more for nomenclatural reasons (because there were no other available names at species rank) rather than for taxonomic ones. It is relevant to note that the ITS1+2 sequences of the two species are substantially different (FIGS. 1–4).

On the other hand, Agaricus moelleri published in 1976 by Wasser is here considered as a synonym with A. praeclaresquamosus (see below) and in accordance with the Saint Louis Code, the former takes priority over the latter. According to the original description of A. moelleri, it differs from A. meleagris (=A. praeclaresquamosus) by its smaller spores and cheilo-cystidia and coloration of the pileus surface. The microscopical data (spores 5 x 3 µm and cheilocystidia 15–18 x 8–12 µm) and pileus color (nigro-squamosus as in Tricholoma terreum (Fr.) Kumm.) given by Wasser for his A. moelleri are the same as those given by Møller for his A. meleagris var. terricolor, this latter being cited by Wasser as synonymous with A. moelleri. In turn, Wasser’s measurements of the spores and cheilocystidia are almost identical to those given by Schäffer (1925) and Møller (1952) for Psalliota meleagris (4–5 x 3 µm and 10–20 x 10–14 µm respectively). [Schäffer did not describe cheylocystidia in the original diagnosis of Psalliota. He pointed out that the lamellar edge is covered by a reddish-brown substance and its microscopical examination is very difficult. For this reason we take the measurement of the cheilocystidia from Møller, who notes that his description of P. meleagris was based on fruit bodies gathered with Schäffer in 1937.] With relation to the pileus surface, its color varies from brown-greyish, through greyish more or less darker to almost black, this trait being very variable and considered by us without taxonomic value. For these reasons, A. moelleri is considered here, as in Rauschert (1992) and Nauta (2001), a synonym with P. meleagris (= A. praeclaresquamosus).

Agaricus parvitigrinus Guinberteau & Callac 2005. Mycologia. in press.

Key synonyms.. none

Characteristic ITS polymorphism.. tggctTtcagg @ 49; gaagcAggtca @ 137; tgggc(-4-)tatgc @ 296–299; ttggtGttccg @ 448

Discussion.. Agaricus parvitigrinus was recently described for specimens collected under Robinia, Sambucus, and Euonymus in calcareous sandy soil in south-western France. This species, which resembles a small form of A. moelleri, is characterized by its slender silhouette, its small size [maximum cap diameter: 5 (6) cm], and its dark grayish squamules that are quite distinct from the white background of the cap. Phylogenetically, it is not close to A. moelleri but belongs to a group of small to medium sized species [maximum diameter: 9 cm] that includes A. californicus, A. laskibarii, and A. xanthodermulus (also recently described). Agaricus parvitigrinus is probably a rare species; however, we recently collected it in different new sites, always in small stands of semi-open middle-aged mixed hardwood forests.

Agaricus phaeolepidotus (F.H. Møller) F.H. Møller 1952. Friesia 4:204. ["1951"]

= Psalliota phaeolepidota F.H. Møller 1952. Friesia 4:170. ["1951"] [basion.]

= Agaricus perdicinus Pilát 1953. Sborn. Nár. Mus. v Praze, ada B, Pír Vdy, 9B(2):24

–Agaricus meleagris var. perdicinus Pilát 1951. Sborn. Nár. Mus. v Praze, ada B, Pír Vdy, 7B(1):108 [nom. inval., nom. nud.; art. 36.1 without Latin description or diagnosis]

Characteristic ITS polymorphisms.. gctggTtctta @ 47; gtatt(-1-)gaggaGgtggtcagTctGtcagc @ 128–149; ygaaaAcgctg @ 191; ctcagCgttca @ 650; tcgtcCtcattttGcTggaca @ 669–679

Discussion.. Agaricus phaeolepidotus is a medium to large species (cap diameter: 6–12 cm), growing in or on the edges of mixed frondose woods. It is a widespread (Austria, British Isles, Czech Republic, Denmark, France, Hungary, Israel, Italy, Spain, The Netherlands, Ukrainia, Uzbekistan, and Belgium) but rare species, based on literature reports. It can be distinguished from the related species A. freirei by its characteristic traits: faint yellowish discoloring of the flesh of the stipe followed by a pale pinkish discoloring mainly at the periphery of the scratched area; cap color varying from pale hazel brown to pale flesh grey brown; annulus pendant, membranous, double, on the underside breaking up into more or less white triangular squamae with brown apex, arranged as a cogwheel farther from the margin; relatively slender habit (as in A. moelleri) with frequent white rhizomorphs; numerous sphero-pedunculate, globose or clavate-piriform cheilocystidia. Agaricus phaeolepidotus forms, with A. freirei and A. hondensis, a basal clade in most analyses, otherwise an isolated clade, within Xanthodermatei.

Agaricus placomyces Peck 1878. New York State Mus. Ann. Rep. 29:40.

Key synonyms.. none.

Characteristic ITS polymorphisms.. ttttcCatagt @ 29; aactgCttgcg @ 558

Discussion.. Agaricus placomyces is a fairly common woodland member of section Xanthodermatei in eastern North America, at least north of Florida. It tends to have a somewhat narrowly conic pileus when young, and a very elastic annulus that can remain attached to the pileus margin even when the pileus is almost fully expanded (to plane), by which point the undersurface of the annulus is ragged and "often studded with drops of a dark-colored fluid" (Peck). These two features, and its slender stature, are generally reliable and distinguish it from other similar species in North America, and from A. moelleri in Europe. Agaricus placomyces is not well documented outside of eastern North America, although the name has been (mis)applied by various authors. See also comments on A. sp. ML5 below.

We (LAP) have studied the type looking for cheilocystidia and found none, only some small clavate or pyriform elements from the edge that could be inconspicuous cheilocystidia. However, an examination (LAP) of specimen RWK 1961 has demonstrated clear and conspicuous cheilocystidia. Agaricus moelleri also has clear and conspicuous cheilocystidia. Therefore both species are rather similar, macro- and microscopically. The understandable historical difficulties in distinguishing A. placomyces from similar species are best remedied by comparing DNA sequences and several traits of their annuli.

The collection RWK 1959, collected in grass near spruce, had four ITS1+2 sequence heteromorphisms, while collection RWK 1918, collected under spruce 20 m away, had none. In our view these admittedly few data suggest that distinct populations of A. placomyces may have re-integrated following a period of isolation and divergence. This phenomenon would also agree with our impression of considerable recent speciation within the group of closely related species, and may partly explain why ‘transitional forms’ are referred to often in the taxonomy of this group.

The excellent photos in Krieger (1936), apparently of material from NY, are of a robust mushroom more reminiscent of A. iodosmus than A. placomyces.

Agaricus pocillator Murrill 1941. Mycologia 33: 446.

Key synonyms.. none.

Characteristic ITS polymorphisms.. gaatcCtgttt @ 225; cctgt(-1-)agagt @ 237; gctca(-1-)gtttt @ 246; ataccTtataa @ 263; tctct(-2-)gctct @ 345–346; attttCtgaat @ 691

Discussion.. Agaricus pocillator is a gracile woodland species of the southeastern United States; Gainsville, FL is the type locality. It is distinguishable by an enlarged, usually cup-like bulb at the base of the stipe: a remarkable and apparently reliable feature. It was described as having a truncate-convex to plane pileus; specimens with fully convex pilei are also known. We attempted to obtain amplifiable DNA from several collections at TENN; unfortunately this was unsuccessful. Presently only the sequence in GenBank (Johnson & Vilgalys 1998) is available. Agaricus pocillator is a relative of A. tollocanensis.

Herbarium material from four southeastern states had generally similar spore sizes: (4.6–) 4.9 (–5.1) x (3.1–) 3.5 (–3.8) µm. These numbers are slightly larger than those in Murrill’s protologue, while the length is somewhat shorter than in A. placomyces (4.8–6.0 µm). Murrill’s collection TENN 16051, from a lawn (an atypical habitat) in Gainesville, FL, also had unusually large spores (5.5–5.9 x 4.1–4.4 µm) and must be considered to be a doubtful determination. Murrill described A. pocillator as "excellent for the table" but as it belongs to section Xanthodermatei we cannot recommend it.

Agaricus pseudopratensis (Bohus) Wasser 1976 Uk-rayins’k. Bot. Zhurn., 33(3):250.

Psalliota pseudopratensis Bohus 1939 Borbasia 1:114. [basion.]

= Agaricus pseudopratensis var. niveus Bohus 1980. Ann. Hist.-Nat. Mus. Natl. Hung. 72:94.

–Agaricus pseudopratensis (Bohus) Bohus 1971. Ann. Hist.-Nat. Mus. Natl. Hung. 63:81. [nom. inval.; art. 33.3 basionym not indicated and reference omitted]

Characteristic ITS polymorphisms.. gggtaTattga @ 125; tcagcAtt(-2-)gctgg @ 155–158; tcattTcatta @ 253; tggccCcttgctTaaggt @ 545–552; cgtct(-1-)cagtg @ 670

Discussion.. Agaricus pseudopratensis, described from material collected in a meadow in Hungary, Fánien Valley, Vértes mountains, is a rare and little known species which usually occurs on sandy soil in coastal dunes, meadows or grassy places, and under Pinus sp., rarely on disturbed or nitrified soils under fruit trees or Pinus spp. in Southern Europe and principally in the Mediterranean area. This species can be distinguished within section Xantho-dermatei by its often squat habit, its simple annulus thickened at the margin with a double edge, and its context at first white, then pale yellow, which after 2–3 min becomes rosaceous and finally, after 5 min, pale reddish to reddish brown in the lower half of the stipe, and also by its claviform cheilocystidia. The pileus surface coloration is very variable: it can be completely white, ochraceous, greyish-brown or greyish or brown, and when colored it is more or less dissociated in triangular radially arranged squamae. Initially, Bohus (1939) did not describe a phenolic odor nor yellow discoloration of the context, but later he described in more detail a yellow discoloration and a slightly phenolic odor at the stipe base for this species (Bohus 1971).

The stature is a variable characteristic. Slender, erect field collections are known. When squat basidiomata from Greece or France were cultured, slender, erect basidiomata were produced (Kerrigan unpubl, Callac unpubl). The ITS1+2 sequence is variable at 5 positions, but not according to a pattern suggestive of infraspecific structure.

Agaricus rotalis K.R. Peterson, Desjardin & Hemmes 2000. Sydowia 52(2):244.

Key synonyms.. none.

Characteristic ITS polymorphisms.. tcagcTtttat @ 145

Discussion.. Agaricus rotalis was described from material collected in the Hawaiian Islands, USA, in soil, duff, woodchips, and on rotting logs under trees, in Common Ironwood Coastal Forest, Alien Wet Forest, and lowland urban areas. This species is macro- and microscopically very similar to Agaricus endoxanthus (see comments under that species) but can be separated from the latter by 3–5 nucleotide differences in the two ITS sequences.

Agaricus tollocanensis Callac & Mata 2005. Doc. Mycol. 33(132): 31–35. ["2004"]

Key synonyms.. none.

Characteristic ITS polymorphisms.. ctagaCgggtt @ 33; ccactGactgg @ 548; gggatCgctct @ 638; attttAtgaat @ 691

Discussion.. Agaricus tollocanensis was recently described for specimens collected in Mexico State, Mexico. It is a sister species of A. pocillator, which is known in the southeastern USA. Both species are characterized by an abruptly bulbous stipe (which is cup-like in A. pocillator) but they differ in the width of their stipe, which is particularly slender in A. pocillator, and by their spores, which are larger in A. tollocanensis (5.7 x 4.1 µm on average).

Agaricus xanthodermulus Callac & Guinberteau 2005. Mycologia. in press.

Key synonyms.. none.

Characteristic ITS polymorphism.. taaatCatcta @ 616

Discussion.. Agaricus xanthodermulus was recently described for specimens collected in grassland or in semi-open middle-aged mixed forest in southwestern France. This species, which resembles a small form of A. xanthodermus, is characterized by its slender silhouette, its small size [maximum cap diameter: 5 (6) cm], its smooth and generally white cap, and its relatively long spores (6.7–7.8 x 4.5–5.5 µm). Phylogenetically, it is closer to A. laskibarii than to A. xanthodermus. Agaricus laskibarii differs from A. xanthodermulus by its medium size (cap diameter: 5.5 to 9 cm) and its shorter spores (5.8–7 x 4.2–5 µm). We consider that the photo of A. pseudopratensis var. niveus 544 on page 192 in Arrillaga (2004) represents specimens of A. xanthodermulus. This is confirmed by their large spores (7.0 x 4.8 µm on average according to our own observations, N = 30). Those specimens were collected in dunes in south-western France. Agaricus pseudopratensis differs from A. xanthodermulus by its smaller spores and by its less slender silhouette, which is often squat.

Agaricus xanthodermus Genev. 1876. Bull. Soc. Bot. France 23:32.

Psalliota xanthoderma (Genev.) Richon & Roze 1886. Atlas Champ. 3:53.

= Agaricus xanthodermus var. lepiotoides Maire 1909. Bull. Soc. Mycol. Fr. 24:LVIII. ["1908"].

= Agaricus lepiotoides (Maire) Konrad & Maubl. 1948. Encycl. Mycol. 14:106. [nom. illeg.; art. 53.1 non Agaricus lepiotoides Berk. & Broome 1871. J. Linn. Soc., Bot. 9:551.].

= Agaricus xanthodermus var. griseus (A. Pearson) Bon & Cappelli 1983. Doc. Mycol. 13(52):16.

Characteristic ITS polymorphisms.. none (but see discussion below).

Discussion.. Presently, A. xanthodermus and A. iodosmus are the only two species in section Xanthodermatei that have been demonstrated, based on ITS sequence data, to occur in both North America and Europe. Agaricus xanthodermus can occur in lawns or under trees (e.g. Cupressus or Olea in California) but is seldom if ever found in deep forest. We regard this species as not recommended for consumption, although it may be regularly eaten by some people in or from Eastern Europe and Russia (RWK pers obs). Both Bohus (1974) and Pilát (1951) have eaten it (or ‘forms’ of it) "without any ill effects" (Pilat 1951). However Bohus (1976) later noted that some collections give off a strong carbolic smell (in agreement with our experience with North American and western European material) and termed these "unsuitable for consumption." Pilát also wrote about a toxic variant (Pilát 1951).

Agaricus xanthodermus differs consistently from A. moelleri at two positions in their ITS1+2 sequences (and occasionally at two other positions). In our experience this level of sequence divergence occurs near the ‘species level boundary’ in Agaricus; in other words, the observed sequence divergence could differentiate either infraspecific or specific taxa (Challen et al 2003). However, in Europe these two taxa are approximately sympatric, yet no evidence of hybridization (e.g. allelic heteromorphisms at positions 158 and 236; cf. Kerrigan 2005) is seen. Based on this observation, biologically based reproductive isolation is evident, therefore we believe that recognition of these two taxa at the species level is appropriate.

Agaricus xanthodermus, like A. moelleri, is very sensitive to the environment: it often grows in curved lines in which sporocarps most exposed to the dryness and/or the sunlight can become colored, fissured, squamulose, or excoriated (lepiotoid), while others remain smooth and white. Without dryness, strong sunlight, or age effects, the pileus of A. xanthodermus is generally white and smooth, while the pileus of A. moelleri is colored and squamulose.

A. sp. CA 215 & 216

Characteristic ITS polymorphisms.. aggggGattgc @ 635

Discussion.. This new species, collected under Abies and Fagus in two sites 150 m apart, is morphologically intermediate between A. moelleri and A. iodosmus: the habit resembles A. moelleri in being more slender, but it is conico-convex rather than truncately conical like A. iodosmus. The pileipellis is dark grey bistre with small white veil remnants when young; breaking up radially and concentrically into square squamules which remain geometrically organised. The annulus has a very broad upper part like A. moelleri, but it is fleshy like A. iodosmus. The base of the stipe is bulbous-submarginate like A. moelleri. The yellow discoloration is strong chrome yellow as for A. iodosmus. Spores and cheilocystidia are similar to those of A. moelleri. The closest species, based on ITS1+2 sequence data, are A. iodosmus, A. placomyces, A. sp. ML5 and A. sp. RWK 1937.

A. sp. RWK 1937
A. sp. ML5

Characteristic ITS polymorphisms.. none.

Discussion.. At least two apparent phylogenetic entities from California have in the past been treated collectively, and incorrectly, as A. praeclar-esquamosus (Kerrigan 1982, 1986). These mushrooms have a more robust stature than the European A. moelleri (see discussion above), brown to blackish pigmentation of the pileus cuticle, a rubbery but inelastic annulus that dehisces promptly and remains largely intact below, and an intense yellowing reaction upon exposure that is generally restricted to the extreme base of the stipe. They grow in coastal forests, including Redwood forest, most often at the edges of trails and roads. Unfortunately, attempts to obtain amplifiable DNA from RWK’s collections at SFSU were unsuccessful. ITS1+2 sequences from two relevant mushrooms that were recently provided to us indicated the presence of two phylogenetic entities, one (RWK 1937) with affinities to A. iodosmus and the other (ML5) closest to A. placomyces. The sequence of ML5 differs from those of A. placomyces by three characters (and additional heteromorphisms in RWK 1959). Six sequence characters distinguish RWK 1937 from A. iodosmus. Five sequence characters distinguish RWK 1937 from ML5.

More collections of these entities will be needed in order to associate DNA-based characters with what appear likely to be fairly subtle morphological features. Until adequate new field material has enabled further analyses, this group of entities must be considered to be incompletely known. Agaricus placomyces sensu Hotson & Stuntz (1938) is likely to be one such species.

A. sp. RWK 1971 & 1972

Characteristic ITS polymorphisms.. cgcctAtctgg @ 71; ctgga(t/A)atgag @ 166; ctcctGtactt @ 497

Discussion.. A distinctive mushroom, presently known from two collections, has recently been encountered in New Mexico, USA. Based on its ITS1+2 sequence, it forms a clade with A. xanthodermus, A. moelleri and A. menieri. Although it might be mistaken for A. xanthodermus, it lacks most or all of the distinctive yellow staining of the cuticles of that species when bruised; the interior of the stipe base does turn yellow at least transiently when exposed. The cuboidal > truncate > umbonate pileus shape, faintly ruddy cast, and often bulbous stipe base also mark A. sp. RWK 1971/1972. While the ITS1+2 sequences clearly indicate that RWK 1971/1972 is a distinct species, further study is in order before a new taxon is described.

Taxa not included in the phylogenetic analysis – A. sp. RWK 1007

Kerrigan (1982) provided a description of a distinctive member of the section based on a single basidiocarp collected in Santa Cruz Co., California, USA by E. Gerry. Attempts to obtain amplifiable DNA from this basidiocarp (SFSU) were unsuccessful. The intense phenolic odor and intense yellow discoloration of the entire lower half of the stipe suggest a relationship to A. iodosmus; however the abundant semi-lustrous "Rood’s brown" to "army brown" (Ridgway 1912) appressed fibrils on the pileus imply that this is a unique phylogenetic entity. Further collections of this rare mushroom are needed.

A. cervinifolius (Zeller) Hotson & Stuntz 1938. Mycologia 30:211.

= Psalliota cervinfolia Zeller 1933. Mycologia 25:388. [basion.]

The type of A. cervinifolius (Peoria district, Linn Co., Oreg. [/] Nov. 20, 1932 [/] Coll. S. M. Zeller [/] Ex. Herb. Oregon State College No. 5589 [NY]) agrees with A. californicus microscopically. Spores from the type measure 5.8–5.9 x 4.2–4.3 µm (RWK). Macroscopically, although similar, P. cervinifolia as described is larger and more robust than A. californicus, somewhat like A. hondensis (Zeller 1933). The pileus cuticle and flesh become yellowish when bruised, much like A. xanthodermus. The odor and taste were reported as pleasant (!) (Zeller 1933). Clearly this taxon merits further study.

Other published European and North American taxa. – Attempts to amplify DNA from the type collections of Agaricus xanthodermus var. macrosporus Aparici & Mahiques (1996), and Agaricus praeclaresquamosus var. macrosporus Aparici & Mahiques (1996), proved unsuccessful. Agaricus phaeolepidotus var. minimus Raithelhuber (1988), Agaricus placomyces var. microsporus A. H. Smith (1940) and A. placomyces var. flavescens Thiers (1959) are apparently rare species of uncertain affinities which will need further study when opportunity permits. Psalliota infida Alessio (1975) should also be evaluated.

Excluded taxa. —. Agaricus bresadolanus Bohus 1969. Ann. Hist.-Nat. Mus. Natl. Hung. 61:154. ["bresadolianus"].

= Agaricus campestris var. pratensis ß radicatus, Vittad. 1832. Descr. Fung. Mang.:42.

= Agaricus campestris var. radicatus (Vittad.) Sacc. 1916. Fl. Ital. Crypt. Hymeniales:800. ["radicata"].

= Psalliota campestris var. radicata (Vittad.) Bres.1931. Icon. Mycol. 17:827.

= Agaricus radicatus (Vittad.) Romagn. 1937. Bull. Soc. Mycol. France 53:129. [nom. illeg.; art. 53.1 non Agaricus radicatus Relh. 1786. Fl. Cantab. suppl. 28 : Fr. 1821. Syst. Mycol. 1:118., qui est Xerula radicata (Relh. : Fr.) Dörfelt 1975. Veröff. Mus. Stradt Gera, Naturwiss. Reihe 2–3:67.].

= Agaricus romagnesii Wasser 1977. Ukrayins’k. Bot. Zhurn. 34(3):305.

Discussion.
This common species is widespread in Europe and according to Wasser (2002)