The filamentous fungus Podospora anserina has been used as a model organism for more than 100 years and has proved to be an invaluable resource in numerous areas of research. Throughout this period, P. anserina has been embroiled in a number of taxonomic controversies regarding the proper name under which it should be called. The most recent taxonomic treatment proposed to change the name of this important species to Triangularia anserina. The results of past name changes of this species indicate that the broader research community is unlikely to accept this change, which will lead to nomenclatural instability and confusion in literature. Here, we review the phylogeny of the species closely related to P. anserina and provide evidence that currently available marker information is insufficient to resolve the relationships amongst many of the lineages. We argue that it is not only premature to propose a new name for P. anserina based on current data, but also that every effort should be made to retain P. anserina as the current name to ensure stability and to minimise confusion in scientific literature. Therefore, we synonymise Triangularia with Podospora and suggest that either the type species of Podospora be moved to P. anserina from P. fimiseda or that all species within the Podosporaceae be placed in the genus Podospora.

phylogenetics, Podospora, Podosporaceae, taxonomy

Podospora anserina is a model filamentous fungus that has been at the forefront of molecular biology and genetics for over 100 years (Silar 2020). It has been instrumental in numerous important biological breakthroughs, such as the discovery of eukaryotic plasmids and prions and has been monumental in furthering the fields surrounding aging/senescence, meiotic drive, allorecognition (known as heterokaryon incompatibility in fungi), sexual reproduction and genome defence (Esser 1974; Saupe et al. 2000; Saupe 2007; Silar 2013, 2020; Grognet et al. 2019; Vogan et al. 2019). Along with its role in basic research, P. anserina has also caught the attention of industry, where it is used as a source of enzymes that play various roles in the degradation of plant material (Couturier et al. 2016). More recently, P. anserina has burst into the genomics era with one of the first published fungal genomes (Fitzpatrick et al. 2006), released in 2008 (Espagne et al. 2008). In the last year, 10 additional chromosome level assemblies of P. anserina have been released in concert with the genome of the closely-related species, P. comata and P. pauciseta (Silar et al. 2019; Vogan et al. 2019). Future projects will expand on this role even further. Wageningen University hosts a collection of strains isolated from the same locale, spanning 30 years of sampling (van der Gaag et al. 1998, 2000; Vogan et al. 2020), which have now all been sequenced and chromosome level assemblies (in preparation) have been produced for the remaining four species of the Podospora anserina species complex (Boucher et al. 2017). Therefore, the role of P. anserina will continue to be central to many fields in biology and, indeed, likely see use in new fields as new data become public.

The taxonomic history of Podospora anserina has been a long and complex one (reviewed in Silar 2020). Podospora is a member of the Sordariales and has been traditionally grouped within the family Lasiosphaeriaceae, which itself is an artificial assemblage of genera whose main diagnostic is that they do not belong to the Sordariaceae (Lundqvist 1972). Species within the Lasiosphaeriaceae were divided into genera, based primarily on ascospore morphology, but molecular phylogenies have revealed that these characters do not represent synapomorphies and that most of the genera are polyphyletic (Huhndorf et al. 2004; Miller and Huhndorf 2005). A broad phylogenetic treatment of coprophilous Lasiosphaeriaceae defined four separate clades, with species of Podospora represented in all clades, exemplifying the lack of informative morphology amongst these fungi (Kruys et al. 2015). Moreover, the taxon P. anserina itself has survived multiple attempts to rename it in the past, which were unsuccessful in part due to how deeply ingrained the name is in the genetics and molecular biology research community (Boucher et al. 2017; Silar 2020). Recently, this taxonomic mess was stumbled upon by researchers attempting to clean up the distantly-related genus Thielavia (Wang et al. 2019). The authors defined the clade containing the type species of Podospora (P. fimiseda, incorrectly referred to as P. fimicola in Wang et al. 2019) as the Podosporaceae (formerly Lasiosphaeriaceae IV) based on a four-marker phylogeny and further divided this clade into three genera: Podospora, Triangularia and Cladorrhinum. As their analyses suggested that P. anserina is more closely related to the type species of Triangularia (T. bambusae), they proposed the new combination, Triangularia anserina (Wang et al. 2019).

It is the opinion of the authors here that the taxonomic change of P. anserina is both premature and against the ideals of the International Code of Nomenclature for algae, fungi and plants (ICN), as stated in the preamble (Turland et al. 2018). Foremost, it is unlikely that T. anserina will be adopted by many of the researchers that rely on it as a model organism, leading to instability of the name. Furthermore, the phylogeny upon which the change was based has a sparse sampling of the diversity of the family and used only four markers. Here, we demonstrate that there is a lack of information amongst the markers currently sequenced in this group and argue that more data are needed before formal taxonomic changes are made for Podospora. Ultimately, the best solution for taxonomic stability in the Podosporaceae will be to change the type species of Podospora from P. fimiseda, which was conserved in 1972 (Nicolson et al. 1984), to P. anserina and to only define new genera once more sequence data are available, likely in the form of whole genome sequences.

Our complete dataset contains 107 taxa and 5895 sites, of which 2110 are variable and 1654 are parsimony informative (Suppl. material 5: Table S2: Supplementary_Table2_Markers). However, combined datasets have a considerable amount of missing data due to the sparse availability of markers for most species. In agreement with previous studies, Maximum Likelihood analyses of all markers reveal that three well-supported clades are resolved within the family, referred to here as Clade A, Clade B and Clade C (Fig. 1; see also Suppl. material 1: Fig. S1 ITSLSU.min0.45–250, Suppl. material 2: Fig. S2 Btub1_and_2.min0.45–250 and Suppl. material 3: Fig. S3 rpb2.min0.45–250). The exception is Btub1 and Btub2, alone or combined, which tend to place members of the outgroup within the ingroup (Suppl. material 2: Fig. S2 Btub1_and_2.min0.45–250). The taxon Cladorrhinum brunnescens appears to represent a distinct lineage within the family, but this finding is only based on the rDNA markers, as no other markers are available for this taxon. None of the main clades of the combined dataset shows monophyly for any genera included in the analyses and, for some species, like Arnium olerum, representative strains appear to be highly divergent. The focal species of this paper, P. anserina, falls within Clade A (Fig. 2), whereas the type species of Podospora, P. fimiseda, is in Clade C (Fig. 3). While each clade itself is generally well supported for individual markers and combinations of markers, the relationships between the clades are not (Suppl. material 1: Fig. S1 ITSLSU.min0.45–250, Suppl. material 2: Fig. S2 Btub1_and_2.min0.45–250 and Suppl. material 3: Fig. S3 rpb2.min0.45–250). The combined analysis of all markers shows support for the sister relationship of Clades A and B, as reported previously by Wang et al. (2019), but this topology seems to be driven exclusively by the rpb2 marker (Fig. 1). A concatenated analysis of all markers, excluding rpb2, recovers a sister relationship between Clades A and C instead, albeit poorly supported. Except for rpb2, individual markers have generally low ΔGLS values (that is, the difference in likelihood between competing topologies is small), indicating that they have relatively low support for either potential sister relationship (AB or AC). By contrast, rpb2 is strongly biased towards the AB hypothesis (Fig. 4A). Notably, the majority of sites for most markers have higher support for the AC relationship, including rpb2 (Fig. 4B). This suggests that the AC clustering is often favoured by any given site, but only weakly (i.e. the difference in likelihood is very small). Although less frequent, the sites in rpb2 that do support the AB relationship have a large likelihood difference between topologies and likely drive the overall positive ΔGLS value of this gene. Thus, the strong degree of conflict between markers for this internode seems to be driven by a single gene with strong phylogenetic signal (rpb2) and for several other markers without sufficient phylogenetic signal.

Figure 1. 

Schematic phylogenetic relationships of the main clades within the Podosporaceae based on Maximum Likelihood analyses of concatenated markers. The three main clades (A, B and C) are strongly supported (bootstrap support values next to relevant branches), but their particular relationship changes depending on the presence of the rpb2 marker. Branches proportional to the scale bar (nucleotide substitutions per site).

Figure 2. 

Maximum Likelihood phylogeny of the concatenated analysis of ITS, LSU, Btub and rpb2 for the Podosporaceae, with an emphasis on Clade A. Type strains are indicated with a bold T and those of the focal species Podospora anserina and Triangularia bambusae are highlighted with coloured boxes. Bootstrap support values are depicted next to their respective branches, but values corresponding to nearly identical sequences are removed for clarity. Branches are proportional to the scale bar (nucleotide substitutions per site).

Figure 3. 

Maximum Likelihood phylogeny of the concatenated analysis of ITS, LSU, Btub and rpb2 for the Podosporaceae, with an emphasis on the clades B and C. Type strains are indicated with a bold T and that of the focal species Podospora fimiseda is highlighted with a coloured box. Bootstrap support values are depicted next to their respective branches, but values corresponding to nearly identical sequences are removed for clarity. Branches are proportional to the scale bar (nucleotide substitutions per site).

Figure 4. 

Phylogenetic signal in the available molecular markers for the relationship between clade A and either clade B or C of the Podosporaceae A differences in the gene-wise log-likelihood scores (ΔGLS) for each marker, where 0 implies equal support for either of the two alternative sister relationships (A and B or A and C), positive values mean higher support for A and B and negative values higher support for A and C B proportion of sites that support each of the two sister relationships within each marker.

With the widespread use of molecular markers to determine the phylogenetic relationships among species, numerous important fungal groups have faced taxonomic challenges including Cryptococcus (Hagen et al. 2015; Kwon-Chung et al. 2017), Zymoseptoria (Quaedvlieg et al. 2011), Fusarium (Geiser et al. 2013), Magnaporthe (Zhang et al. 2016) and Ustilago (McTaggart et al. 2016; Thines 2016), many of which remain unresolved. Fungi are, of course, not the only group whose phylogenetic and taxonomic concepts are in conflict; the model insect genus, Drosophila, is still embroiled in a nomenclatural controversy that has yet to be satisfactorily resolved (O’Grady and Markow 2009; Johnson 2018). In the case presented before us, the issue is, in fact, not so complex as most of these other examples. The taxonomic assignments amongst species and strains of the Sordariales has long been recognised as a difficult problem, with various authors seeking to provide some clarity (Huhndorf et al. 2004; Miller and Huhndorf 2005; Félix 2015; Kruys et al. 2015). It is clear that previous use of morphological characters to designate genera has failed to resolve monophyletic groups, as all genera represented here are not only distributed amongst the three clades within the Podosporaceae, but can also be found in the much more distantly-related clades (Lasiosphaeriacceae I–III sensu Kruys et al. (2015)). It is thus apparent that the way forward is to define the genera based on molecular phylogenies.

While previous phylogenetic analyses on the Sordariales in general have been informative, the lack of resolution remains a pervasive issue (e.g. Kruys et al. 2015). Our results suggest that the molecular markers, typically used for the study of this group, have a relatively low phylogenetic signal for a number of key internodes. Within Podosporaceae, in particular, the relationship between the clades containing P. anserina and P. fimiseda (clades A and C) is crucial to decide on an optimal naming scheme that minimises taxonomic and practical conflict. Additionally, throughout the phylogeny, there are a number of strains that have been assigned to different species and genera, but likely belong to the same species (e.g. Thielavia hyalocarpa and Zopfiella leucotricha strain CBS389.84 are identical for ITS and LSU). Additionally, there appears to be undiscovered sexual states of the anamorphic Cladorrhinum species, like C. foecundissimum with the A. olerum strain CBS120012. Taxonomic re-assignments of these groups should be undertaken; however, without a strong phylogenetic backbone, based on multiple genes and an expanded taxonomic sampling, it seems premature to propose nomenclatural changes.

In recent years, a number of authors have established the use of time-calibrated phylogenies to define ranks from genus up to class for various groups of fungi, although this approach has not been without controversy (Lücking 2019). For the Sordariomycetes, intervals of 150–250 MYA for orders and 50–150 MYA for families have been recommended (Hyde et al. 2017). Both the Sordariales and the Podosporaceae agree well with these values with estimated divergence times of 109.69 MYA and < 76.58 MYA, respectively (Lutzoni et al. 2018), although a comprehensive investigation of divergence amongst the Lasiosphaeriaceae has yet to be conducted. The use of time-calibrated trees to define genera is less common, but a divergence time of ~30 MYA has been suggested previously (Divakar et al. 2017). In this regard, it would be appropriate to define the three clades presented herein as genera as Wang et al. (2019) have done, based on divergence times between Lasiosphaeria ovina and Anopodium ampullaceum (Hyde et al. 2017), which show similar phylogenetic distances between each other as the three clades of Podosporaceae (Félix 2015). However, time-calibrated phylogenies need not and often should not, set the standard for taxonomic delimitation. One telling example comes from the field of Fusarium research. When the ‘one name one fungus’ edict came into effect, it threatened to divide a cosmopolitan group of plant pathogenic fungi that are united in their vegetative morphology and pathophysiology into several genera. To combat this issue, the field pushed for the usage of Fusarium for the entire group, despite considerable phylogenetic distance (Geiser et al. 2013). As a result, Fusarium encompasses species which diverged more than ~70 MYA (Lutzoni et al. 2018), yet this move ensured the nomenclatural stability of the organisms in question.

When proposing new combinations, one should always ensure to make decisions that will cause the least amount of confusion in literature. In this case, it is clear that, for this goal, the name P. anserina should be preserved. Google Scholar returns ~11500 hits to the search query Podospora anserina, yet only ~2110 hits for Triangularia, the majority of which are due to the use of the word “triangularia” in Latin and have no relation to the genus. There are currently 62967 sequences in Genbank with Podospora anserina in the title, while only 76 contain Triangularia in the title and lastly, the English Wikipedia page for Podospora anserina has had 13897 page views from July 2015 to May 2020, while the English Wikipedia Triangularia page has had only 705 views over the same period. The best possible way forward to prevent the re-naming of P. anserina or the subsequent instability it will cause in literature is to transfer the type of Podospora from P. fimiseda to P. anserina, despite P. fimiseda having been conserved over Schizothecium fimicola Corda (Proposal 119). Unfortunately, this process can take many years of debate and the re-assignment of P. anserina to Triangularia already threatens a peaceful transition. At the very least, if P. anserina needs to be assigned to another genus, it should be the type species of that genus in order to prevent further potential nomenclatural changes. Thus, we propose, for the interim, to synonymise Triangularia with Podospora until a more satisfactory resolution can be made. Once more data are available, it will hopefully be possible to resolve the relationships amongst the three clades. If, in the end, Clade A and B are found to be sisters, then it would require that Cladorrhinum sensu Wang et al. (2019) (Clade B here) be synonymised with Podospora as well. Alternatively, if Clade A and C are sisters, then Podospora could be restricted to these two clades and fewer taxonomic changes would be required. As we ultimately aim to move the type species of Podospora to P. anserina, we will refrain from making any new combinations at the moment.

In the previous example with Fusarium, a divergent group of fungi were classified under one name precisely because the researchers in that field desired unity. In the case of Podospora, the only factor necessitating that disparate species fall under one genus is the need to operate within the confines of the ICN. The ultimate goal of the code is to provide taxonomic stability and conformity to the organisms it covers. The nature of studying microscopic fungi has resulted in numerous names with dubious origins and, while obvious fixes are sometimes evident, they are not always possible to enact according to the ICN. It is understandable why the current code is as rigid as it is, but the current editions have seen it become more flexible, which has been advantageous to many fields seeking to solidify tumultuous taxonomy. In the future, we hope that additional data and a permissive code will allow us to enshrine the name Podospora anserina indefinitely, settling over a century of nomenclatural friction between taxonomists and other researchers.

We would like to thank Lymari Ruiz for their assistance in generating the sequence data, to Iker Irisarri for his guidance with the phylogenetic analysis, and Martin Ryberg and Mats Thulin for their advice on taxonomic issues. We would like to thank the Swedish Research Council and Formas for research funds. We would also like to acknowledge past and present researchers who have devoted much of their time to the study of Podospora anserina to elevate it to the status of model organism.