1OTU numbering system of Nagy et al. (this issue of New Phytologist, pp. 789–794).
Where are all the undocumented fungal species? A study of Mortierella demonstrates the need for sequence-based classification
Article first published online: 20 JUL 2011
© 2011 The Authors. New Phytologist © 2011 New Phytologist Trust
Volume 191, Issue 3, pages 592–596, August 2011
How to Cite
Hibbett, D. and Glotzer, D. (2011), Where are all the undocumented fungal species? A study of Mortierella demonstrates the need for sequence-based classification. New Phytologist, 191: 592–596. doi: 10.1111/j.1469-8137.2011.03819.x
- Issue published online: 20 JUL 2011
- Article first published online: 20 JUL 2011
- molecular ecology;
How many different kinds of fungi inhabit the earth? Hawksworth’s (2001) estimate of 1.5 million extant species is widely cited, but many other figures have been proposed, from a ‘lower limit’ of 712 000 species (Schmit & Mueller, 2007) to over 5 million species (O’Brien et al., 2005). While there is great disparity among these estimates, they all suggest that the c. 100 000 species that have been described (Kirk et al., 2008) represent just a small part of the actual diversity of fungi. It follows that most unidentifiable environmental sequences probably represent new species, which has led to proposals for sequence-based taxonomy (Hibbett et al., 2011). In this issue of New Phytologist, Nagy et al. (pp. 789–794) assess the gap between the described and sequenced dimensions of fungal diversity, focusing on the zygomycete genus Mortierella (and the related mitosporic taxa Umbelopsis, Gamsiella and Dissophora).
‘As long as the locality and substrate data are consistent, these modern sequences could be applied to older names. Doing so would honour past authors and forge a link to the literature …’
Nagy et al. downloaded 832 mortierelloid internal transcribed spacer (ITS) sequences from GenBank (http://www.ncbi.nlm.nih.gov/genbank/) and combined them with a newly constructed reference dataset of 102 sequences from type or authentic strains, representing 78 described species (including synonyms). Using a 97% sequence similarity cut-off, they grouped the sequences into 92 molecular operational taxonomic units (MOTUs). Fifty-two MOTUs contained one or more sequences from the reference dataset, but 40 MOTUs could not be identified. Observing a roughly linear relationship between the number of reference sequences that were sampled and MOTUs that could be identified, Nagy et al. predicted that all of the MOTUs could be identified if only 49 more species could be added to the reference dataset, bringing it up to 127 species in total. Because this is close to the number of known species of mortierelloid fungi, Nagy et al. concluded that ‘the vast majority of Mortierella and Umbelopsis species have already been described,’ and therefore ‘most “unidentifiable” environmental sequences in fact represent species already described by taxonomists’.
In the absence of sequences from type strains of all the described species of mortierelloid fungi we cannot determine whether the unnamed MOTUs reported by Nagy et al. represent new or rediscovered taxa. Instead, we focus here on the relative rate of description of new species of mortierelloid (and other) fungi and the discovery of MOTUs, and we consider the implications for the practice of fungal taxonomy in the age of high-throughput environmental sequencing.
Nomenclature of mortierelloid fungi is complex. According to the Zygomycetes website of G. L. Benny (http://zygomycetes.org) there are 109 nonredundant names in Mortierella, Umbelopsis, Dissophora and Gamsiella, but there are 244 names within these genera listed on MycoBank (http://mycobank.org), which were published between 1863 and 2004. The 78 species that Nagy et al. recorded in their reference dataset correspond to c. 61 species sensu Benny (i.e. 55% of the recognized species). Descriptions of mortierelloid fungi peaked in the 1940s and again in the 1960s and 1970s (Fig. 1), with the work of G. Linneman, T. V. Khalabuda, W. Gams, and others (see http://zygomycetes.org or http://mycobank.org for literature).
Information on the composition of the MOTUs, including GenBank accession numbers, was kindly provided by László G. Nagy. Using a combination of automated text-based searches and manual inspection of GenBank records, we attempted to divide the ITS sequences into environmental sequences (those obtained by direct amplification from DNA extracted from environmental substrates) and culture-based sequences. The sources of some sequences are ambiguous, but we estimate that 509 of the ITS sequences (55%) are of environmental origin. Environmental sequences occur in 23 of the identified MOTUs and 30 of the unnamed MOTUs. Twenty-four of the unnamed MOTUs are composed solely of environmental sequences. Thirty-seven of the unnamed MOTUs, including all of the purely environmental MOTUs, were discovered since 2005 (Table 1).
|MOTU1||Number of sequences2||Number of studies||Oldest sequence||Newest sequence|
|60||56||8||49||15||2000 AJ271630 Mortierella alpina||2011 GU055570 Uncultured Mortierella|
|59||40||19||12||12||2005 AJ879650 Uncultured Mortierellaceae||2010 FN598954 Mortierella alpina|
|3||27||19||8||13||2005 AY627838 Epacris pulchella root associated fungus||2011 FJ554273 Uncultured Mucorales (aff. Umbelopsis) envseq|
|46||8||4||4||4||2005 AY970014 Uncultured fungus||2011 FJ554273 Uncultured Mucorales (aff. Umbelopsis)|
|76||6||5||0||5||2007 EF434039 Uncultured fungus||2009 FJ237094 Uncultured fungus|
|69||5||5||0||3||2005 AY969763 Uncultured fungus||2011 FJ554400 Uncultured eukaryote envseq|
|6||4||1||3||4||2007 EU076951 Umbelopsis sp.||2010 GQ241270 Umbelopsis sp.|
|8||4||0||4||4||2004 AY618256 Umbelopsis isabellina||2010 AY805542 Umbelopsis isabellina|
|15||4||0||4||2||2003 AY172097 Cf. Verticillium sp.||2009 FJ161928 Mortierella elongata|
|71||4||4||0||2||2008 EF521230 Uncultured fungus||2010 GU559105 Uncultured fungus|
|81||4||4||0||4||2008 AM260850 Uncultured fungus||2010 GU328567 Uncultured Basidiomycota|
|78||3||3||0||3||2006 Zygomycete sp.||2008 EU292438 Uncultured fungus|
|84||3||1||2||2||2006 AJ878781 Mortierella macrocystis||2011 AB476532 Uncultured fungus|
|4||2||2||0||2||2008 DQ273434 Uncultured Mucorales||2011 FJ553764 Uncultured Mucorales (aff. Umbelopsis)|
|9||2||0||2||2||2005 AY781211 Umbelopsis isabellina||2010 GU062324 Umbelopsis sp.|
|21||2||2||0||2||2005 AY970096 Uncultured ascomycete||2009 FN397150 Uncultured fungus|
|24||2||1||0||2||2010 GU366718 Uncultured fungus||‘Alaska B4’|
|49||2||2||0||2||2005 AY970066 Uncultured fungus||2010 EU917096 Uncultured fungus|
|50||2||2||0||2||2005 AY969665 Uncultured fungus||2010 AB542110 Mortierella sp.|
|73||2||2||0||1||2006 EF152542 Fungal sp.||2006 EF152543 Fungal sp.|
|12||1||0||1||1||2006 DQ118996 Mucor racemosus||–|
|19||1||0||1||1||2007 EU076935 Mortierella sp.||–|
|20||1||1||0||1||2010 EU917106 Uncultured fungus||–|
|28||1||1||0||1||2005 AY969671 Uncultured fungus||–|
|29||1||1||0||1||2006 DQ309131 Uncultured fungus||–|
|33||1||0||1||1||2009 AB476423 Mortierella umbellate||–|
|45||1||0||1||1||2006 AJ878779 Mortierella horticola||–|
|51||1||1||0||1||2005 AY969588 Uncultured fungus||–|
|52||1||1||0||1||2008 DQ420859 Uncultured soil fungus||–|
|56||1||0||1||1||2010 GU396234 Mortierella sp.||–|
|61||1||0||1||1||2007 EF192184 Mortierella alpina||–|
|62||1||1||0||1||2010 EU917123 Uncultured fungus||–|
|68||1||1||0||1||2007 AM901914 Uncultured zygomycete||–|
|72||1||1||0||1||2008 EF521225 Uncultured fungus||–|
|74||1||1||0||1||2005 AY627790 Uncultured fungus||–|
|82||1||1||0||1||2007 EF040832 Uncultured fungus||–|
|83||1||1||0||1||2009 EU554792 Uncultured fungus||–|
|85||1||1||0||1||2008 EF521232 Uncultured fungus||–|
|89||1||1||0||1||2007 EF619911 Uncultured zygomycete||–|
|91||1||0||1||1||2008 EU877747 Zygomycete||–|
In a previous commentary (Hibbett et al., 2009), we asked if taxonomists or molecular ecologists are currently leading the way in the discovery of fungal biodiversity. If the environmental MOTUs are new species, then molecular ecology is clearly the major arena of contemporary species discovery (Fig. 1). Even if one assumes that all the unnamed MOTUs represent previously described species, as suggested by Nagy et al., then it is still impressive that in only 6 yr molecular ecologists have rediscovered more than half of the species of mortierelloid fungi that were described over 140 yr by taxonomists (corresponding to the 53 MOTUs that contain one or more environmental sequences). Either way, the situation in the mortierelloid fungi prompts us to ask what can be done to reconcile the unnamed MOTUs and the described species that lack sequences.
The response of Nagy et al. was to call for expanded sequencing of type cultures and specimens. We agree that it is important to enhance identification resources by sequencing reference collections, but we are not optimistic that type sequences can be obtained for all fungal taxa. Mortierelloid fungi may represent a best-case scenario for building a reference dataset, because they are easily isolated and grow well in culture. Nevertheless, Nagy et al. were able to locate type or authentic material for only about half of the described species, despite an intensive effort. At this point, it may be more productive to seek neotypes or epitypes for the taxa that lack sequence data. (Neotypes are used when original type material is not extant, whereas epitypes are used when type material is available but is ‘demonstrably ambiguous and cannot be critically identified for purposes of the precise application of the name of a taxon’, International Code of Botanical Nomenclature (ICBN) Art. 9.7, http://ibot.sav.sk/icbn/main.htm.)
Sixteen of the unnamed MOTUs include sequences that appear to be derived from cultures, which might be appropriate for use as neotypes or epitypes, as long as the observable features of the new cultures are consistent with the descriptions of the taxa that lack sequence data. However, most of the unnamed MOTUs are not associated with cultures or other physical specimens, and they cannot serve as neotypes under the current ICBN (at least as it is typically interpreted). To connect these sequences to published taxon names it will be necessary to develop protocols for neotypification or epitypification using environmental sequences (including rules for priority). Criteria for sequence-based taxonomy are still emerging, but there is a general sense that singleton sequences are not sufficiently reliable to warrant naming. Fortunately, 20 of the unnamed MOTUs have two or more sequences, and in all but one case the sequences are from different studies (Table 1). As long as the locality and substrate data are consistent, these modern sequences could be applied to older names. Doing so would honour past authors and forge a link to the literature, while providing identification resources for fungal molecular systematists and ecologists.
Are fungal taxonomists running out of species to describe?
Nagy et al. were careful to note that the results they obtained for mortierelloid fungi might not be replicated in other groups, but they nonetheless suggested that ‘Mortierella may be used as an example of general trends in fungi’, implying that the actual diversity of fungi is much closer to the 100 000 described species than has previously been thought. If the global catalogue of fungal diversity is nearing completion, then we would expect the rate of species description to be declining. To address whether this is happening, we plotted the total number of new species and infraspecific taxa in all groups of fungi described annually from 1950 to 2010, based on records in the Index of Fungi (http://www.cabi.org/). Over that interval, an average of 1499 new species and infraspecific taxa have been described each year (91 451 total). The slope of the regression (new taxa vs year) is not significantly different from zero, suggesting that we are nowhere near saturation (Fig. 2a). This analysis assumes constancy of sampling effort, which is very difficult to measure. As a crude proxy for ‘effort’, we used the size of the annual subscription list of the journal Mycologia from 1967 to 2005 (from the published statements of ownership; data from 2006 to 2010 were omitted out of concern that recent declines in Mycologia subscriptions could skew the results). We suggest that these numbers might reflect the level of activity in the fungal taxonomy community. We then calculated corrected taxon counts by dividing the number of taxa by the number of subscriptions to Mycologia. Using the corrected counts, there is a significant positive slope, suggesting that the rate of taxon description per unit of effort measured this way is actually increasing (Fig. 2b). These results, together with the flood of unidentifiable sequences from environmental studies (Hibbett et al., 2011), and discoveries of cryptic species-rich clades, such as the Rozellida/Cryptomycota clade (Lara et al., 2010; Jones et al., 2011) or ‘soil clone group 1’ (Porter et al., 2008), suggest that fungal biodiversity remains a wide-open frontier, even if some groups (possibly including mortierelloid fungi) have been better documented than others.
The authors are grateful to László G. Nagy for providing information about the MOTUs, and Paul Kirk for providing data on new taxa recorded in the Index of Fungi. This work was supported by AFTOL (NSF award DEB-0732968).
- 2001. The magnitude of fungal diversity: the 1.5 million species estimate revisited. Mycological Research 105: 1422–1432. .
- 2011. Progress in molecular and morphological taxon discovery in Fungi and options for formal classification of environmental sequences. Fungal Biology Reviews 25: 38–47. , , , , , .
- 2009. Fungal ecology catches fire. New Phytologist 184: 279–282. , , .
- 2011. Discovery of novel intermediate forms redefines the fungal tree of life. Nature 474: 200–203. , , , , , , .
- 2008. Dictionary of the fungi, 10th edn . Wallingford, UK: CABI. , , , eds .
- 2010. The environmental clade LKM11 and Rozella form the deepest branching clade of fungi. Protist 161: 116–121. , , .
- 2011. Where is the unseen fungal diversity hidden? A study of Mortierella reveals a large contribution of reference collections to the identification of fungal environmental sequences. New Phytologist 191: 789–794. , , , , , .
- 2005. Fungal community analysis by large-scale sequencing of environmental samples. Applied and Environmental Microbiology 71: 5544–5550. , , , , .
- 2008. Widespread occurrence and phylogenetic placement of a soil clone group adds a prominent new branch to the fungal tree of life. Molecular Phylogenetics and Evolution 46: 635–644. , , , , , , , .
- 2007. An estimate of the lower limit of global fungal diversity. Biodiversity and Conservation 16: 99–111. , .