The fungal species: a matter of definition

The debate over how species should be defined (the ‘species concept’) has been going on so long that it seems to be an endless biological and philosophical tradition; and it’s not likely soon to arrive at a universal definition of ‘species’ that meets with widespread agreement. Indeed, as recently as 1997 an author could write a paper that discussed 22 different species concepts in current circulation (Mayden, 1997). We do not wish to contribute to this debate, but we do need to give you a brief outline of the most commonly applied species concepts in fungal taxonomy.

The morphological species concept. Historically, biologists, including mycologists, first recognized species based on morphological similarity; this creates what are known as morphospecies. The difficulty with this is in finding characters that convincingly define the boundaries of a species (you need to know the boundaries so that the next specimen can either be placed within or outside that species). If you doubt the difficulty, look around at other people and think what morphological features you might use for describing the human species. Body morphology? Skin colour? Eye colour? Hair distribution? Face shape? You use all of these to identify individuals, but they are poor candidates to circumscribe the species. In fungi, there are not only fewer morphological characters to use, but they are also very variable, and the natural variation of characters that can in addition be influenced by environmental variation is difficult to measure. Add to this the fact that similar morphologies can be arrived at by very different evolutionary routes (convergent evolution) and you’ll not be surprised to learn that most biologists and mycologists in particular, believe that the morphological species concept is the least satisfactory. Unfortunately, the foundation for fungal (particularly mushroom) classification was established by Elias Fries between about 1820 and 1875, with emphasis on morphological features, such as gills, pores, and spore colour. This was practical for identification, but it obscured the phylogenetic origins of many of these features.

The biological species concept. The dominant idea in biology (especially animal biology) is the biological species concept in which a species is defined as an interbreeding population that is somehow reproductively isolated from other populations. This looks simple and obvious but is also flawed; and especially so for fungi. The first problem is to identify the reproductive barrier(s) between populations made up of morphologically similar individuals. The concept is not even applicable to homothallic and asexual organisms, which excludes about 20% of fungi; nor is it applicable to the large number of organisms that cannot be cultivated, because it depends on mating tests being carried out on specimens raised in the laboratory. The biological species concept also has difficulty coping with populations that are geographically isolated (or rather, the meaning of geographic isolation is difficult to apply consistently). Populations that do not interbreed because of geographic isolation will evolve independently and may well diverge sufficiently to become different species on other criteria, and yet still interbreed successfully when brought together artificially if they retain common ancestral sexual characteristics. The underlying question, though, is what constitutes an effective reproductive barrier. Geography, even global geography, doesn’t help with fungi because within the group there are many organisms that produce spores that can glide through the atmosphere across and between continents, ignoring oceans and mountain ranges. At the other extreme, there are microscopic fungi that may be located in such restricted habitats that they are effectively isolated from their relatives just a few metres away. Overall, there are too many severe restrictions on application of the biological species concept widely in fungi for it to be a serious contender for that elusive universal definition.

Ecological and physiological species concepts. It seems a reasonable assumption that parasitic or symbiotic fungi have at least some measure of host-specificity and that such a fungal species might be defined on the basis of habitat and/or host relationships. Equally reasonable are the expectations that ecological adaptations influence fungal speciation, and that physiological features contributing to adaptation to habitat and/or host could characterize a fungal species. In effect, the habitat and/or host relationship is being cast in the role of reproductive isolation mechanism. An ecological and/or physiological species concept has been used for a long time with plant pathogenic fungi. The concept chiefly differentiates species by their ecological niche and the constraints on their evolution that determine their and maintenance and reproduction in that niche (and the niche may be a particular species of host plant, or even a specific cultivar of a host species). Again, it sounds reasonable, but there are problems. In particular, we are largely ignorant of the exact physiological/biochemical/genetical nature of whatever it is that determines substrate specificity and host specificity, so it is little better than using morphological characters. Furthermore, application of the concept is severely limited in practice. A plant pathologist might be able to characterise a fungus of interest from its host spectrum, or a medical mycologist from its serotype, but there is a vast range of other fungi for which this approach simply fails. It is another concept that cannot offer the universal definition that would be most valuable.

Evolutionary/phylogenetic species concept. Definitions based on molecular analyses seem likely to be the most promising. Their foundation is the ancestry of the species. They envisage a species as being a monophyletic group of organisms sharing molecular characters that derive from a common ancestor. This, and the fact that the operation is done with the DNA sequence itself, is inherently satisfying. Furthermore, the concept does not have any obvious built-in exclusions or limitations. Particularly important to fungi is the fact that the analysis can be applied to asexual organisms; anamorphic and teleomorphic stages can be covered by a single species concept. Nor is there any shortage of characters – fungal genomes are big enough to provide more than enough sequences for this to be a definition of universal applicability. Clearly, multiple gene analyses will provide the best understanding and this highlights the current limitation, which is that we don’t have enough information yet.

However, a new strategy based on the sequencing of standardised genomic fragments (DNA barcoding) was established early in the 21st century (Costa & Carvalho, 2007; Hollingsworth, 2007). The DNA barcode approach involves the use of a standard DNA region that is specific for a taxonomic group. The DNA marker you choose is amplified and sequenced. The sequencer generates a ‘barcode’ because it detects the four bases that make up the sequence using fluorescent dyes and the sequencer’s colorimeter reports the sequence by printing appropriately coloured peaks (or bars) that designate the sequence. If the sequence contains a run of the same base, then the peak (or bar) is broad; where the sequence has a single base, the peak (or bar) is narrow. So, just as the unique pattern of bars in a universal product code (UPC) in the supermarket identifies each consumer product, a ‘DNA barcode’ is a unique pattern of DNA sequence that identifies each living thing (providing you’ve made the right choice of DNA marker!). Botanists tend to use a portion of the chloroplast DNA of plants, those interested in animals can use mitochondrial cytochrome c oxidase subunit I or 16S rRNA. Schoch et al. (2012) identified several DNA regions as promising universal barcodes for fungi and found that the nuclear ribosomal RNA internal transcribed spacer (ITS) region exhibited the highest probability of correct identification for a range of fungal lineages. Since then, the ITS region has been accepted as the standard barcode marker for fungi, although there are indications that this sequence is not always useful in identifying genera and a careful, informed selection must be made (Badotti et al., 2017; Truong et al., 2017). A specific example is a study of which ITS primers give optimal amplification in species of Tuber that showed that two fungal specific primer pairs (ITS1f and ITS6) have a mismatch in one base pair in the target sequence of Tuber aestivum, thus resulting in poor or no amplification success (Unuk Nahberger et al., 2020). Another word of warning is that different analytical platforms yield different outcomes from the same data sets. The time required for computation differs, but more importantly, the quality of error filtering and hence the quality of the output largely depends on the platform used. Therefore, taxonomic assignments derived from ITS data need to be validated by other methodologies (Anslan et al., 2018). However, as techniques improve it may be only a matter of effort and time for this approach to be applied effectively to a good range of fungal species (Castaño et al., 2020; Runnel et al., 2021). Runnel et al. (2021) have shown that fungal diversity assessments based on DNA metabarcoding in wood of fallen Norway spruce trunks were the same as in sporophore surveys, leading to the same conservation management conclusions.

Different species concepts = different outcomes. The use of different methods of defining species inevitably results in recognition of different things. The indications are that application of the phylogenetic species concept will lead to the recognition of far more species than have already been recognized using morphological, biological or physiological species concepts.

A recent study across all major groups of eukaryotes examining the effect of applying the phylogenetic species concept to organisms previously well-classified using other criteria showed an average 48% increase in the number of species documented. The increase tends to be greater in fungi. In most traditional fungal groups, a reclassification using molecular methods together with the phylogenetic species concept leads to a 2 to 4 times increase in the number of species recognised.

This means more names and more taxonomy, which is not greeted very sympathetically by those who are not taxonomists. But just think what biological truth is being revealed here. If molecular sequence analysis characteristically reveals 2-4 times more species than we know about using non-molecular methods, then the extent of biodiversity and the species richness of every ecosystem are 2-4 times greater than we currently imagine. And that’s just for ‘species’ we know about!