Fungal Taxonomy III:
The Euagarics

© Peter Werner
Original publication: Mycena News, March 2004

In my last two articles, I've discussed how our ideas about the relationships between fungi are being radically revised based upon advances in molecular biology and computer-aided phylogenetic analysis. So far, I've discussed discoveries in the taxonomy of "lower" fungi, as well as in ascomycetes, russulales, boletales, and other groups. I now want to focus on the taxonomy of the euagarics, that is, the group that the majority of gilled mushrooms fall into.As I mentioned in my previous article, current evidence infers that the agaric-forming fungi evolved at least three times independently. Lentinoid agarics like Neolentinus ponderosus and Neolentinus lepidius (both previously placed in Lentinus) represent an independent evolution of agarics from polypores and similarly, the Russulaceae represent an independent evolution of agarics from woody resupinate fungi. Several families of agarics, the Paxillaceae and the Gomphideaceae, are closer to boletes then they are to other agarics. However, its not yet clear whether both groups share an agaric ancestor and boletes are a derived condition, or whether agarics in the bolete clade evolved independently from boletoid ancestors.The majority of agarics belong to the euagaric group and share common descent from a single ancestor. The relationships within this group have begun to be examined over the last decade with some rather surprising results. Recently, a group of some 14 mycologists fronted by Jean-Marc Moncalvo carried out a molecular analysis of about 700 species of euagarics to reveal which groups these species would congregate into. This analysis revealed 117 "clades" (distinct monophyletic groups) of euagarics. While further research will undoubtedly reveal affinities between clades that were not revealed in this analysis, it is quite clear that many of these 117 clades have as strong a claim to being to being "families" as the traditional Singerian families. The division of the euagarics into 11 families clearly does not begin to adequately describe the complex of evolutionary relationships within the euagarics.

In the past, spore color was treated as a characteristic of fundamental importance in defining agaric families. However, a pattern that clearly emerges from Moncalvo's study and other molecular studies is that, although spore color is somewhat conserved within euagaric clades, in the overall evolution of agarics, spore color has shifted frequently, hence, its use as a method of dividing the agarics into fundamental groups is not warranted.

The last several "generations" of amateur and professional mycologists have been thoroughly schooled in the Singerian scheme of classifying agarics, and some of the new classifications that are emerging may seem confounding at first, however, in most cases these groupings do make sense in the light of the shared morphologies between members of these taxa.

Click to Enlarge
A small section of the Euagaric "tree".
(from Moncalvo, et al. One hundred and seventeen clades of euagarics. Molecular Phylogenetics and Evolution 23: 357-400.)

Take, for example, our newly expanded concept of the Agaricaceae. The Moncalvo study (as well as several prior independent molecular studies) have revealed a strong affinity between Agaricus and the Lepiotaceae, and even hinted that Agaricus is simply a specialized dark-spored group of within the macrolepiotas. Additionally, Coprinus comatus and its close relatives have recently been found not to be closely related to the other coprinoid fungi (which have now been consigned to the genera Coprinellus, Coprinopsis, and Parasola), instead falling within the expanded Agaricaceae. Most surprisingly, the Lycoperdaceae (the puffballs) have been found to be related to the previously mentioned taxa as well, and hence constitute a highly-specialized gastroid line within the Agaricaceae.

As for the remaining coprinoid agarics, molecular studies have confirmed the affinity between several genera that had constituted the former Coprinaceae. Coprinellus, Coprinopsis, and Psathyrella are clearly related and are now referred to as the family Psathyrellaceae. Panaeolus, however, was found not to be closely related to this group, and instead was found to be very close to Bolbitius and Conocybe, forming the basis for a redefined Bolbiteaceae. The status of the remaining bolbitioid genus, Agrocybe, is uncertain - it seems to form its own separate clade, which may or may not be close the Bolbiteaceae.

Moncalvo's study has also found that the Strophariaceae seem to fall into several distinct clades that may or may not be closely related. One clade contains the core strophariod fungi, including Stropharia, Hypholoma, most Pholiota, and the bluing Psilocybe. Another clade contains the non-bluing Psilocybe, Melanotus, and Kuehneromyces. If further study confirms that the non-bluing and bluing Psilocybe are, in fact, not directly related, it will be split into two genera, with the genus name Psilocybe probably going to the non-bluing species, as the type species for Psilocybe is apparently Psilocybe montana. (Since chemical names don't change with biological nomenclature, a defining character of Psilocybe would then be its lack of psilocybin.) There are also several subgenera of Pholiota, plus several of species of Stropharia and Hypholoma, that fall outside of these clades entirely and their relationship to the other stropharioid agarics is uncertain.

A close relationship has also been demonstrated between Pluteus and the Amanitaceae, which forms the basis for a newly expanded Pluteaceae. Oddly, Volvariella, an agaric with clear morphological affinities to both Pluteus and Amanita, was found in the Moncalvo study not to fall into the same clade as these genera, and that in fact, its relationship is quite distant. This provides a case where the molecular and morphological evidence seem to be in conflict and further study is clearly necessary.

Studies on the Hygrophoraceae are contradictory, with the Moncalvo study suggesting that the Hygrophoraceae as conceived in the Singerian system is a clear monophyletic clade, and other studies suggesting that the Hygrocybe is not directly related to the rest of the hygrophoroid fungi, but that a monophyletic clade is formed by Hygrocybe, Omphalina, perhaps Xeromphalina, and the remaining hygrophoroid agarics.

The Entolomataceae seem to be an exception to the above pattern, in that it actually holds together as a distinct group centered on its traditionally-defined genera without substantial revision. As would be expected, the Tricholomataceae and the Cortinariaceae do not hold together at all - these have always clearly been "wastebasket taxa"; a catch-all for all white- and rusty-brown-spored agarics that couldn't be placed in any other family. The majority of the 100+ clades of agarics found in the Moncalvo study are segregates from these two families. These family names will be retained, but only in much more limited sense for fungi that are closely related to Tricholoma and Cortinarius, respectively. Reclassification of some of these segregates has already been established by prior morphological work, and several of these even have "family" names that are increasingly coming into use, for example, the Pleurotaceae (Pleurotus and Hohenbuehelia), the Marasmiaceae (Marasmius, Tetrapyrgos, and allies), and the Crepidotaceae (Crepidotus, Simocybe, and allies).

As we've seen over the course of the last three articles on this topic, molecular biology is reshaping our understanding of relationships between the fungi at all levels, from phyla down to species. I'm often asked why we should accept such revisions, especially where they contradict established morphological classifications. My answer is that its best to accept any classification system, new or old, only in a provisional way. It should be pointed out that many of our "well-established" morphological classifications really aren't that well-established at all, and that molecular studies give us much-needed additional data to use in our classification systems. A system of classification is based (hopefully) on the best data available at the time, and science is always going to find new data. Our knowledge of evolutionary history and actual evolutionary history are two different things, and the former is always only an approximation of the latter, one that hopefully, over time, improves.

Further reading: