Phylogenetic Considerations

Placement of Laccaria in Ordinal and Family Treatments
	A robust hypothesis of phylogenetic relationships for the Agaricales has not been developed and several classifications have been advanced that reflect differences in views of relationships among genera in the group (e.g., Kühner, 1980; Jülich, 1981; Singer, 1986). Kühner (1980, 1984) proposed dividing the Agaricales into several small orders that he felt were easier to define and had sharper boundaries than traditionally recognized (e.g., Singer, 1975). He treated Laccaria, along with its gasteroid counterpart Hydnangium, as a separate family named the Hydnangiaceae within his order Tricholomatales (Kühner, 1980, 1984) (Table 2). Hydnangium is the sister taxon to Laccaria and the Tricholomataceae the sister taxon to the Hydnangiaceae in this classification. Jülich (1981) also recognized the Tricholomatales as a separate order but treated Laccaria in his monotypic family Laccariaceae separate from, but closely related to, the Hydnangiaceae (Table 2). These two families are among the 22 families that he recognized for the order (Jülich, 1981). Unfortunately, Jülich was not explicit on intra-ordinal relationships except to say that he did not recognize close relationships among many of these fungi. Thus it is not clear what he thought was the sister group to the Laccariaceae and Hydnangiaceae. Singer (1986) maintained that separation of the Agaricales into a number of smaller orders is not warranted. Within his classification, Singer (1986) placed Laccaria in a separate subtribe (Laccariinae) in Tribus Tricholomateae of the Tricholomataceae (Table 2). Singer maintained that Hydnangium, as well as all other Gasteromycetes, is sufficiently distinct to prevent it from being treated in the Agaricales. Subtribe Clitocybinae is the sister taxon to Laccaria and the Termitomyceteae is the sister taxon to the Tricholomateae in this classification. Molecular data obtained to date do not contain information with which the Agaricales sensu Singer can be divided into smaller orders. Rehner et al. (1990) reported that their sequence data of mitochondrial and nuclear ribosomal RNA genes do not resolve phylogenetic relationships within the Agaricales except to support the separation of the Boletales and the Russulales from the rest of the agarics. Lacking supportive evidence from molecular analyses and heeding Singer's (1986) arguments for the use of caution regarding the multiplication and upgrading of higher taxa, I treat Laccaria, along with Hydnangium and Podohydnangium (see below), within the Tricholomataceae. It was beyond the scope of this study to try to resolve phylogenetic relationships within the Tricholomataceae. Until these relationships have been rigorously analysed, I do not recognize infrafamiliar taxa in the Tricholomataceae. However, I accept Singer's (1986) and Kühner's (1980) view that Laccaria has closer affinities with genera in Singer's Tricholomateae and Kühner's Tricholomataceae than other genera in the Tricholomataceae sensu Singer (1986).
	Table 2. Placement of Laccaria in Kühner's (1980), Jülich's (1981) and Singer's (1986) classifications of Agaricoid Hymenomycetes. Kühner, 1980 TRICHOLOMATALES Tricholomataceae Clitocybeae Gerroneama, Omphalina, Arrhenia, Cantharellula, Clitocybe, Lepista, Ripartites, Armillaria, Lyophylleae Biannularieae Tricholomeae Tricolomopsis, Leucopaxillus, Tricholoma, Pseudobaeospora, Melanoleuca Cystodermateae Hydnangiaceae Laccaria, Hydnangium Rhodotaceae Pleurotaceae Marasmiaceae Hygrophoraceae Amanitaceae Jülich, 1981 TRICHOLOMATALES Hygrophoraceae Laccariaceae Laccaria Hydnangiaceae Hydnangium Fayodiaceae Collybiaceae Macrocystidiaceae Xerulaceae Gigaspermaceae Armillariaceae Biannulariaceae Tricholomataceae Leucopaxillaceae Lyophyllaceae Marasmiaceae Amparoinaceae Mycenaceae Favolaschiaceae Nyctalidaceae Cyphellopsidaceae Lachnellaceae Resupinataceae Rhodotaceae Singer, 1986 AGARICALES Tricholomataceae Lyophylleae Termitomyceteae Tricholomateae Laccariinae Laccaria Clitocybinae Clitocybe, Lepista, Tricholomopsis Tricholomatinae Tricholoma Omphalinae Armillariella, Arthrosporella, Lulesia, Arrhenia, Leptoglossum, Omphalina, Gerronema, Callistosporium, Pleurocollybia, Lactocollybia, Macrocystidia, Fissolimbus, Asproinocybe Leucopaxilleae Biannularieae Collybieae Resupinateae Panelleae Marasmieae Myceneae Pseudohiatulaeae Rhodoteae
Relationship of Laccaria to Hydnangium and Podohydnangium
	Podohydnangium must be considered along with Hydnangium in discussions of the relationship of Laccaria to putative gasteroid relatives. Podohydnangium is a monotypic genus that differs from Hydnangium by having a distinct stipe-columella and, therefore, appears intermediate between Laccaria and Hydnangium (Beaton et al., 1984). Podohydnangium was not included in Kühner's (1980) or Jülich's (1981) treatments since it was first described in 1984 and was excluded from Singer's (1986) classification for the same reason that he excluded Hydnangium; uncertainty of affinity. These three classifications discussed above treat the relationship of Laccaria to Hydnangium and Podohydnangium differently (Table 2). This is because of fundamental differences in opinion regarding the relationship of gasteroid genera to their agaric counterparts. Kühner (1980) and Jülich (1981) both treated certain gasteroid taxa within the Agaricales, incorporating them into the classification with their putative sister taxa. Singer (1986), on the other hand, maintained that the relationship of these fungi to epigeous agarics is not sufficiently resolved to justify incorporating them into the Agaricales. While Laccaria, Podohydnangium and Hydnangium differ drastically in macromorphology, they share several presumably derived micromorphological character states; identical basidiospore ornamentation and, at least in Laccaria and Hydnangium (Podohydnangium has not been examined), multinucleate basidiospores. Laccaria is unique among epigeous agarics in that the conic echinulae, diagnostic features of the genus, are formed by microtubials that run perpendicular to the epispore (Besson and Kühner, 1971; Kühner, 1980; v. Hofsten and Mueller, unpublished). SEM micrographs of basidiospores from several Hydnangium and Podohydnangium taxa have documented the similarity of shape of the basidiospore ornamentation between these taxa and those of Laccaria (Pegler and Young, 1979; Beaton et al., 1984; Castellano et al., 1989) and unpublished TEM data obtained by v. Hofsten (Institute of Physiological Botany, University of Uppsala, Uppsala, Sweden) document that the basidiospore wall ultrastructure of Hydnangium is similar to that found in Laccaria; the echinulae are composed of microtubials that run perpendicular to the epispore. All examined taxa of Laccaria as well as Hydnangium carneum have multinucleate basidiospores (Table 1; Kühner, 1980; Tommerup et al., 1991). The three genera also share several pleisiomorphies such as having abundant clamp connections; nonamyloid, acyanophilic basidiospores; and the lack of a heteromerous trama (Pegler and Young, 1979; Beaton et al., 1984). Finally, at least some species of Hydnangium appear similar in color to orange-brown Laccaria. Differences between the genera are statismosporic basidiospores that are orthotropic in development in Hydnangium and Podohydnangium versus ballistosporic basidiospores that are heterotropic in development in Laccaria. But, as pointed out by several authors (e.g., Pegler and Young, 1979; Beaton et al., 1984), Hydnangium is not closely related to any of the genera such as Octavianina O. Kuntze formerly treated in the polyphyletic Hymenogastrales sensu Singer and Smith (1960) and others. While Laccaria, Hydnangium and Podohydnangium appear to form a monophyletic group, it is currently not possible to undertake a rigorous analysis of the relationship of these three genera to each other as no detailed systematic work has been carried out on either Hydnangium or Podohydnangium and species circumscriptions, composition, and relationships are still uncertain in these two genera. Castellano and Trappe (1990) accepted 23 names in Hydnangium in their bibliographic survey of Australian gasteroid fungi. Numerous systematic problems remain in the group, however, and some of these taxa probably belong in other genera. Until intra- and intergeneric relationships within this group are resolved, I choose to treat the taxa in this group as three separate genera (Laccaria, Hydnangium, Podohydnangium) in the Tricholomataceae sensu Singer (1986) amended to include Hydnangium and Podohydnangium. The inability to resolve infrafamiliar relationships within the family precludes recognizing the clade composed of these genera in a formal classification (see above).
Previously Published Infrageneric Classifications Proposed for Laccaria
	Three primary infrageneric classifications have been proposed for Laccaria (Bon, 1983; Singer, 1986; Ballero and Contu, 1989) (Table 3). Moser (1983), Clémençon (1984) and others have published variations of these classifications that do not differ significantly from those presented in Table 3. The primary difference between Bon's and Ballero and Contu's classification versus that of Singer's is the rank that they recognized subgeneric taxa. Singer (1986) used the term stirps (a term not recognized by the International Code of Botanical Nomenclature, Greuter et al., 1988) rather than section to reflect the small hiatus between the subgeneric groups that he accepted. Bon (1983) and Ballero and Contu (1989) recognized three sections with Section Laccaria (incorrectly named Laccata by Bon, ICBN Art. 22.1, Greuter et al., 1988) further divided into several subgroups. All three classifications recognized a separate group composed of L. maritima and L. trullissata and a group composed of L. amethystina and other taxa with violet to purple basidiomata (Table 3). Bon (1983) and Ballero and Contu (1989) each recognized the species that have bisterigmate basidia as a separate subgroup. Singer (1986) included these taxa among species with tetrasterigmate basidia. He recognized Stirps Galerinoides and Purpureobadia on differences in basidioma colors (Singer, 1986). Other differences between these three classifications (Table 3) are due primarily to either varying interpretation of taxa or inclusion of different taxa.
	Table 3. Primary infrageneric classifications proposed for Laccaria. Names in parentheses are the correct name for the taxon.
	M. Bon, 1983 Sect. Maritimae Bon L. trullissata L. maritima Sect. Amethystinae Bon L. amethystina L. bicolor L. purpureobadia Sect. Laccata Stirps Ohiensis L. ohiensis (=L. impolita) L. tortilis L. altaica (=L. pumila) L. striatula (=?) L. lateritia (=L. fraterna) Stirps Laccata L. laccata L. affinis (=L. laccata var. pallidifolia) Stirps Tetraspora L. proxima (=?) L. tetraspora (=L. ohiensis) R. Singer, 1986 Stirps Trullissata L. trullissata L. maritima Stirps Amethystina L. ochropurpurea L. bicolor L. calospora (=L. amethystina) L. amethystina (= L. amethysteo occidentalis) L. lilacina Stirps Laccata L. laccata L. farinaceae (=L. trichodermophora) L. proximella L. proxima L. tetraspora (=L. ohiensis) L. montana L. altaica (=L. pumila) L. echinospora (=L. tortilis) L. fraterna Stirps Galerinoides L. galerinoides L. vinaceoavellanea Stirps Purpureobadia L. purpureobadia Ballero & Contu, 1989 Sect. Maritimae Bon L. trullissata L. maritima Sect. Amethystinae Bon L. amethystea (=L. amethystina) L. calospora (=L. amethystina) L. violaceonigra L. masonii L. bicolor L. bullulifera L. farinaceae (=L. trichodermophora) Sect. Laccaria Subsect. Bisporae Contu L. echinospora (=L. tortilis) L. singeri (=L. impolita) L. lateritia (=L. fraterna) L. pumila Subsect. Laccaria L. purpureobadia L. lutea (=?) L. proxima L. montana L. laccata L. tetraspora (=L. ohiensis) L. affinis (=L. laccata var. pallidifolia)
Cladistic Analyses
	Cladistic analyses were undertaken to resolve infrageneric relationships within Laccaria. The infrageneric classifications of Bon (1983), Singer (1986) and Ballero and Contu (1989) discussed above were based on those authors' interpretations of the evolutionary history of the group. These authors, however, did not provide explicit statements regarding crucial assumptions and decisons used to develop their classifications. Without explicit information on choice and weighting of characters, character state evolution, homoplasy (convergence or parallelisms), etc., it is impossible to make an objective comparison or rigorously choose between these conflicting classifications. While cladistics has been used commonly in studies of plants and animals, such analyses have infrequently been used in fungal systematics. Cladistic methods are based on the theories first expounded by Hennig (1966) which have subsequently been expanded and modified by numerous workers (see reviews in Eldredge and Cracraft, 1980; Nelson and Platnick, 1981; Wiley, 1981; Duncan and Stuessy, 1984; Stuessy, 1990). The theoretical basis for these methods includes the following: 1) phylogenetic relationships are based on the idea of common ancestry (monophyly); 2) evidence for monophyletic groups (all, and only, the descendants of a particular ancestor) can only be determined by the detection of shared derived character states (synapomorphies); and 3) shared primitive character states (symplesiomorphies) do not provide evidence for relationships. Although there are operational and theoretical difficulties involved with any technique, cladistic analyses provide a rigorous method for making hypotheses of phylogenetic relationships. Unfortunately, Laccaria does not lend itself to cladistic analyses. Problems in determining outgroups and using the outgroup criterion for character state polarization are treated in the next subsection. Problems in choosing characters and determining character state homology are also treated below. Care must be used, therefore, when interpreting the following results from the cladistic analyses. They are presented only as a working hypothesis of the evolutionary history of the genus, subject to revision. Discussion of choice of outgroups. Lack of consensus regarding relationships within the Agaricales sensu Singer (1986) prevented me from determining an unequivocal choice to use as an outgroup (see discussion at the beginning of this chapter and data in Tables 2-3). Uncertainty is not limited to the determination of the sister group to Laccaria, but also to the generic composition and circumscriptions of potential outgroups (Kühner, 1980, 1984; Singer, 1986). The classifications presented in Table 2 propose conflicting hypotheses regarding the sister taxon to Laccaria. I used the entire Tricholomateae Singer, minus Laccaria, as the sister group to Laccaria. This group of taxa is roughly equivalent to the Clitocybeae plus Tricholomeae in Kühner's Tricholomataceae (Kühner, 1980, 1984). Cantharellus was used as the sister group to the Tricholomataceae. Although some workers do not derive the Tricholomataceae from a Cantharellus-like ancestor (e.g., Singer, 1986), I find this to be a plausible hypothesis following the arguments of many workers (e.g., Petersen, 1971; Kühner, 1980, 1984; Bigelow, 1982). It was well beyond the scope of this study to try to resolve relationships within the heterogeneous group included in the Tricholomateae. Instead, I divided the group into smaller units based only on characters shared by the outgroup and ingroup. Using this criterion I used the following operational units as outgroups: Cantharellus, Clitocybe, Lepista, Tricholomopsis 1 & 2 (differing by basidiospore shape) [represented by Tricholomopsis flavissima (Smith) Singer and T. rutilians (Schaeff.: Fr.) Singer, respectively], Tricholoma 1 & 2 (differing by basidiospore shape) [represented by Tricholoma aurantium (Fr.) Ricken and T. michiganense Smith, respectively], Omphalinae 1 & 2 (differing by basidiospore shape) [represented by Omphalina hepatica (Fr.) Orton and Leptoglossum rickenii (Hora) Singer, respectively], and Omphalinae 3-5 (differing by the number of sterigmata per basidium and number of nuclei per basidiospore) [represented by Gerroneama chrysophyllum (Fr.) Singer, Omphalina pseudoandrosaceae (Bull.) Moser and O. griseopallida (Desm.) Quél., respectively] (Table 5). For ingroups, I used all of the Laccaria taxa recognized from the continental United States and Canada. I further divided L. laccata into three groups L.lac 1-3 (differing in basidiospore shape) because of the high degree of plasticity observed in this character in this taxon. Extralimital taxa were not included in these analyses because of a lack of information on somatic culture mat morphology and difficulties in interpreting the macromorphology, especially color, for a number of taxa that I know only from the literature and an examination of their type specimens. Another serious problem in determining suitable taxa to use as outgroups to Laccaria is that many of the systematically informative characters in Laccaria do not occur in prospective outgroups (e.g., length and width of basidiospore echinulae and, as coded, basidioma and culture mat pigments). It also was impossible to determine character state homology for characters such as basidioma and basidiospore size between putative outgroups and Laccaria taxa. Because of these problems, only 5 of the 14 characters employed in the final analyses could be coded for both the outgroups and ingroups (Tables 4 and 5). Thus most of the ingroup characters could not be polarized directly by outgroup comparison (Watrous and Wheeler, 1981; Maddison et al., 1984). Three separate sets of analyses were run. First, outgroup and ingroup relationships were examined using the five shared characters (Figure 3). Second, resolution of ingroup relationships was attempted using all available characters for the ingroup without the outgroups. This resulted in the unrooted network presented in Figure 4. Finally, to develop an operational classification, analyses of the combined data set were run to identify functional outgroups within Laccaria (Watrous and Wheeler, 1981), thereby constraining possible tree topologies (Figure 5).
Discussion of characters and character state assignments.
	Table 4 lists the characters and their states used in these analyses. Other characters (e.g., presence or absence of striations, pileus texture, growth rate of cultures on various media, etc.) were employed in preliminary analyses but were subsequently deleted for various reasons including that they were either highly variable within and among taxa or were autapomorphic (varied only in one terminal taxon and thus did not provide information on relationships among taxa). Size characters (nos. 6, 17, 18) were only coded for ingroup taxa because of the impossibility of estimating homology with states in the outgroups. Color characters (nos. 11, 12, 13, 19) were limited to ingroup taxa for the same reason. To my knowledge, the identity and structure of the pigments in Laccaria are unknown (see DISCUSSION OF SYSTEMATIC CHARACTERS). Thus, it is not possible to determine homology of pigments observed in Laccaria with those found in other genera with orange-brown and violet pigments. Lack of knowledge of the pigments in Laccaria causes problems in ingroup as well as outgroup analyses. I have made the assumption that the violet coloration observed in many Laccaria, from the lamellae and stipe basal mycelium in L. bicolor, to the entire basidioma in L. amethystina, is due to the same pigment(s). Information on number of sterigmata per basidium and number of nuclei per basidiospore (characters 1 and 2, respectively; Table 4) for outgroups was obtained from a number of sources including Corner (1966), Kühner (1980, 1984), Bigelow (1982) and Singer (1986).

	Table 4. Characters and their states used in cladistic analyses. 1. Number of sterigmata per basidium. 0 = 4; 1 = 2-3; 2 = 5-8 2. Number of nuclei/basidiospore. 0 = 1; 1 = multi 3. Echinulate basidiospores with echinulae formed by perpendicular microtubials. 0 = absent; 1 = present 4. Mean echinulae length (in µm). 0 = rugulose; 1 = < 0.5; 2 = 0.5-1; 3 = 1-2; 4 = > 2 5. Echinulae base width (in µm). 0 = * 1; 1 = „ 1.2 6. Mean basidiospore length. 0 = short (< 8µm); 1 = moderate (8-10µm); 2 = long (10-13µm); 3 = elongate (> 13.5 µm) 7. Basidiospore shape (). 0 = globose ( = 1-1.05); 1 = subglobose to broadly ellipsoid ( = 1.06-1.23); 2 = ellipsoid ( = 1.24-1.6); 3 = oblong ( = 1.65-2); 4 = cylindrical or fusiform ( > 2) 8. Cheilocystidia shape. 0 = absent or filamentous and not strongly morphologically differentiated; 1 = subclavate to clavate; 2 = very large and inflated 9. Pileus color when young and fresh. 0 = flesh color to orange-brown; 1 = violaceous; 2 = red brown; 3 = violet brown; 4 = ochraceous; 5 = rust 10. Lamellar color when young and fresh. 0 = flesh color; 1 = vinaceous; 2 = violet to purple; 3 = rose-pink 11. Basal mycelium color when young and fresh. 0 = white; 1 = violet 12. Pileus size. 0 = moderate; 1 = small; 2 = large 13. Stature. 0 = moderate; 1 = gracile; 2 = robust 14. Color of somatic culture mat on PDA and MMN. 0 = white to olive brown; 1 = violet Table 5. Taxon by character state matrix used in cladistic analyses. Characters and character states provided in Table 4. Outgroups employed were Cantharellus and species of Singer's (1986) Tricholomataceae excluding Laccaria (Table 2). Canth (Cantharellus) 2 0 0 ? ? 2 0 ? ? ? ? ? ? ? Clito (Clitocybe) 0 0 0 ? ? ? 2 0 ? ? ? ? ? ? Lepist (Lepista) 0 0 0 ? ? ? 2 0 ? ? ? ? ? ? Tmopsis 1 (Tricholomopsis 1) 0 0 0 ? ? ? 0 2 ? ? ? ? ? ? Tmopsis 2 (Tricholomopsis 2) 0 0 0 ? ? ? 2 2 ? ? ? ? ? ? Trich 1 (Tricholoma 1) 0 0 0 ? ? ? 1 0 ? ? ? ? ? ? Trich 2 (Tricholoma 2) 0 0 0 ? ? ? 2 0 ? ? ? ? ? ? Omphal 1 (Omphalinae 1) 0 0 0 ? ? ? 1 0 ? ? ? ? ? ? Omphal 2 (Omphalinae 2) 0 0 0 ? ? ? 2 0 ? ? ? ? ? ? Omphal 3 (Omphalinae 3) 0 1 0 ? ? ? 2 0 ? ? ? ? ? ? Omphal 4 (Omphalinae 4) 1 0 0 ? ? ? 2 0 ? ? ? ? ? ? Omphal 5 (Omphalinae 5) 1 1 0 ? ? ? 2 0 ? ? ? ? ? ? L.a-o (amethysteo-occidentalis) 0 1 1 3 0 1 1 1 1 2 1 2 0 1 L.ame (amethystina) 0 1 1 3 1 1 0 1 1 2 1 0 0 1 L.bic (bicolor) 0 1 1 3 0 0 1 0 0 1 1 0 0 1 L.frat (fraterna) 1 1 1 3 0 2 1 0 5 3 0 0 0 0 L.lac 1 (laccata 1) 0 1 1 3 0 1 0 0 0 0 0 0 0 0 L.lac 2 (laccata 2) 0 1 1 3 0 1 1 0 0 0 0 0 0 0 L.lac 3 (laccata 3) 0 1 1 3 0 1 2 0 0 0 0 0 0 0 L.long (longipes) 0 1 1 3 0 0 1 0 0 0 0 0 0 0 L.mar (maritima) 0 1 1 1 0 3 3 0 2 2 1 0 2 ? L.mont (montana) 0 1 1 3 0 2 1 0 0 0 0 1 1 0 L.nob (nobilis) 0 1 1 3 0 0 1 0 0 1 1 2 2 1 L.obl (oblongospora) 0 1 1 2 0 1 2 0 0 0 1 0 0 1 L.och (ochropurpurea) 0 1 1 3 0 1 0 0 4 2 1 2 2 1 L.ohi (ohiensis) 0 1 1 4 1 1 0 0 0 0 0 1 1 0 L.pro (proxima) 0 1 1 2 0 1 2 0 0 0 0 0 0 0 L.pum (pumila) 1 1 1 3 0 2 1 0 0 0 0 1 1 0 L.stri (striatula) 0 1 1 4 1 1 0 0 0 0 0 0 1 0 L.tort (tortilis) 1 1 1 4 1 2 0 0 0 0 0 1 1 ? L.tric (trichodermophora) 0 1 1 3 0 0 1 0 0 0 1 0 0 1 L.trul (trullissata) 0 1 1 0 0 3 4 0 1 2 1 2 2 1 L.v-b (vinaceobrunnea) 0 1 1 3 0 1 1 1 3 2 1 0 0 1
Preliminary hypothesis of phylogenetic relationships based on cladistic analyses.
	Analyses undertaken for this study were performed using PAUP v. 3.0L (Swofford, 1989) running on a MAC II ci. All multistate characters were interpreted as unordered because I could not make a priori decisions on character state transformation series. The number of taxa included in the analyses precluded the use of exact methods to find the shortest trees so a heuristic method using branch swapping (tree bisection-reconnection) was employed to search for optimum trees. The MULPARS option was invoked to save 300 equally most parsimonious trees. A shorter tree was sometimes located even after 200 trees were saved. Ten replications using random addition sequences were employed to ensure that addition sequence did not impact tree length or topology. Characters and their states employed in the final sets of analyses are listed in Tables 4-5. A strict consensus tree of the 300 most parsimonious trees saved in the analysis restricted to the characters that could be coded for both ingroups and outgroups is presented in Figure 3. This tree has a length of 28 steps with a homoplasy index (HI) of 0.643 and a rescaled consistency index (RC) of 0.211. While ingroup relationships were not resolved in this analysis, Laccaria formed a monophyletic clade apart from the outgroup taxa supported by one synapomorphy, basidiospore echinulae ultrastructure (character 3). Fig. 3. Strict consensus tree resulting from analyses restricted to characters shared by ingroup and outgroup taxa (characters 1-3, 7, and 8 in Table 4). Only characters 3 and 8 were not homoplasious. Length 28, HI=0.643, RC=0.211. Refer to Table 5 for abbreviations.
	Only five characters could be used in this analysis (Table 4, characters 1-3, 7 and 8) and of these only characters 3 (presence or absence of echinulate basidiospores) and 8 (cheilocystidia shape) were not homoplasious. Character 3 is the synapomorphy supporting the Laccaria clade while character 8 occurs in 3 states. Most taxa lack or have filamentous cheilocystidia (state 0) while inflated cheilocystidia can be found in two of the outgroups. Laccaria amethysteo occidentalis, L. amethystina and L. vinaceobrunnea have clavate cheilocystidia. The distribution of multinucleate basidiospores (character 2) along this tree is significant because Kühner (1980, 1984) emphasized the occurrence of multinucleate basidiospores in Laccaria in his rational for recognizing the genus as a family separate from other genera typically treated in the Tricholomataceae (Table 2). While Laccaria is characterized by having multinucleate basidiospores, this character state is also present in two of the outgroup taxa (character 2, Figure 3). Until the phylogeny of the outgroup (Singer's Tricholomataceae excluding Laccaria) is elucidated, it is impossible to tell if the presence of multinucleate basidiospores is sympleisiomorphic or was derived independently in two or more clades within the Tricholomataceae. Figure 4 presents the results of the analysis of ingroup relationships utilizing all of the characters listed in Table 4 (minus characters 2 and 3 since they were uninformative for ingroup comparisons). The strict consensus tree of the 300 most parsimonious trees saved is presented as a unrooted network since no outgroup was used to root the tree. The network (Figure 4) has a length of 51 steps, a consistency index (CI) of 0.549 and a rescaled consistency index (RC) of 0.345. Only select characters were traced onto the network illustrated in Figure 4. All characters are mapped along the tree presented in Figure 5 (analysis of combined outgroup and ingroup data). All characters except for echinulae length (character 4), cheilocystidia (character 8), pileus color (character 9) and color of lamellae (character 10) were homoplasious. Because of this high level of homoplasy, the network is not fully resolved and not robust. The addition or deletion of a character or change in coding of the states of one character had profound impact on network topology. Several subgroups within Laccaria were resolved during these analyses (Figure 4). Laccaria ohiensis, L. striatula and L. tortilis formed a trichotomy separate from the other North American taxa that lack violet pigments (Figure 4). This clade was supported by the presence of strongly echinulate, globose basidiospores (characters 4, 5, 7) and small gracile basidiomata (characters 12 and 13). Laccaria proxima and L. oblongospora formed a clade supported by the presence of finely echinulate basidiospores (state 2, character 4). These two clades along with the other taxa that lack violet basidioma pigments formed an unresolved "bush" separate from taxa with violet basidioma pigments. Resolution was higher within the North American taxa with violet pigments. This grade was supported by the presence of violet mycelium at the stipe base and violet culture mats on PDA and MMN media (characters 11, 14). Both of these states, however, occur also in L. oblongospora which was not placed in this grade. The L. bicolor complex (i.e., L. trichodermophora, L. nobilis and L. bicolor) was fully resolved and formed a sister group to the taxa with violet to purple lamellae. Laccaria trullissata and L. maritima formed a trichotomy with L. ochropurpurea due to their large basidioma size and stature (characters 12 and 13) and similarity in coloration (characters 9-11). The clade comprised of L. amethysteo-occidentalis, L. amethystina and L. vinaceobrunnea was supported by the presence of large cheilocystidia (character 8). The distribution of violet basal mycelium and violet somatic culture mats on PDA and MMN media (state 1 of characters 11 and 14, respectively) along this network causes difficulties in interpreting character state changes throughout Laccaria. Based on the topology of the network, violet pigments occur in two places along the tree; L. oblongospora and the grade that includes L. trichodermophora - L. vinaceobrunnea. It is impossible to determine which is the pleisiomorphic state for these two characters. The occurrence of these states in two areas of the tree could be due either to a parallel gain of this pigment(s), a reversal back to the pleisiomorphic state, or the incorrect placement of L. oblongospora. This is a serious weakness in these results because, as discussed below, the major subgroups identified through these analyses are supported by pigment composition and their distribution. Based on these analyses, bisterigmate basidia (state 1, character 1) arose independently three times within the North American taxa of Laccaria; L. fraterna, L. pumila and L. tortilis. Laccaria laccata was divided into three operational taxa in these analyses because of plasticity of basidiospore shape within individuals referable on other character states to L. laccata. These analyses were uninformative for resolving at what rank to recognize these morphological forms of L. laccata. Combining the data sets (all characters for both ingroup and outgroup taxa) was undertaken to constrain the choice for a function outgroup and provide a operational classification based on a tentative hypothesis of relationships among the ingroup taxa. Because only five characters were shared by the ingroup and outgroup, the results of this analysis provide only one of the possible resolutions of where to root the network presented in Figure 4. Figure 5 shows the results of this analysis. The resulting strict consensus tree of the 300 most parsimonious trees saved was 64 steps long and had a consistency index (CI) of .500 and a rescaled consistency index (RC) of 0.332. Fig. 5. Strict consensus cladogram using all characters listed in Table 4 showing one possible resolution of where to root the network presented in Figure 4. Only characters 4, 8, 9, and 10 were not homoplasious. Refer to Table 5 for abbreviations.
	The unresolved group comprised of the taxa lacking violet basidioma pigments plus L. oblongospora are basal on this tree (Figure 5). The only difference in ingroup topology between this tree and the unrooted network (Figure 4) is that L. oblongospora and L. proxima are not placed together on a separate clade.